ICML 2025 Schedule

Filter Events

SUN 13 JUL

11 a.m.

Registration West

(ends 5:30 PM)

2 p.m.

Expo Talk Panel:

Gen AI Applications in Amazon Pharmacy

(ends 3:00 PM)

Expo Talk Panel:

OmniLong: A Resource-Effective Context Scaling Framework for Multimodal LLM Fine-tuning

(ends 3:00 PM)

Expo Talk Panel:

Thursday Night Football on Prime Video – Broadcast Innovation

(ends 3:00 PM)

3 p.m.

Break:

Coffee Break

(ends 4:00 PM)

4 p.m.

Expo Talk Panel:

A Unified Framework for Generative AI Safety

(ends 5:00 PM)

Expo Talk Panel:

Distributed Computing Architectures as a Solution to AI's Energy Crisis: Empirical Analysis of Decentralized Training

(ends 5:00 PM)

Expo Talk Panel:

JokeEval: Are the Jokes Funny? Review of Computational Evaluation Techniques to improve Joke Generation

(ends 5:00 PM)

5 p.m.

Expo Talk Panel:

Re-Imagine: Symbolic Benchmark Synthesis for Reasoning Evaluation

(ends 6:00 PM)

Expo Talk Panel:

Distillation Scaling Laws

(ends 6:00 PM)

Expo Talk Panel:

Human-Aligned Long-Form Evaluation (HALF-Eval): Framework for Assessing AI-Generated Content

(ends 6:00 PM)

MON 14 JUL

7:30 a.m.

Registration West

(ends 6:00 PM)

Registration East

(ends 3:00 PM)

ICML Lounge Area:

ICML Lounge Area

(ends 6:00 PM)

8 a.m.

Expo Talk Panel:

Situating principles in context for synthetic data

(ends 9:00 AM)

Expo Talk Panel:

Otter: Generating Tests from Issues to Validate SWE Patches

(ends 9:00 AM)

Expo Talk Panel:

The Next Frontier in Enterprise AI: A Vision for Generalist Agents

(ends 9:00 AM)

Expo Talk Panel:

Improving LLM Benchmarks: Making AI Work for Real-World Needs

(ends 9:00 AM)

Expo Talk Panel:

Foundation Models for Automated Trading

(ends 9:00 AM)

9 a.m.

Break:

Coffee Break

(ends 10:00 AM)

9:30 a.m.

Affinity Workshop:

LatinX in AI

(ends 5:30 PM)

Tutorial:

Calibration and Bias in Algorithms, Data, and Models: a tutorial on metrics and plots for measuring calibration, bias, fairness, reliability, and robustness

(ends 12:00 PM)

Tutorial:

Modern Methods in Associative Memory

(ends 12:00 PM)

Tutorial:

CANCELED: Alignment Methods for Large Language Models

(ends 12:00 PM)

Tutorial:

Generative AI Meets Reinforcement Learning

(ends 12:00 PM)

Tutorial:

Harnessing Low Dimensionality in Diffusion Models: From Theory to Practice

(ends 12:00 PM)

Tutorial:

DP-fy your DATA: How to (and why) synthesize Differentially Private Synthetic Data

(ends 11:55 AM)

Expo Talk Panel:

Structured Foundation Models Meets AutoML: Shattering the SOTA with AutoGluon & GraphStorm

(ends 10:30 AM)

10 a.m.

Affinity Workshop:

4th MusIML workshop at ICML’25

(ends 4:00 PM)

noon

Break:

Hosted Grab and Go Lunch

(ends 1:30 PM)

1:30 p.m.

Tutorial:

Game-theoretic Statistics and Sequential Anytime-Valid Inference

(ends 4:00 PM)

Tutorial:

Tutorial on Mechanistic Interpretability for Language Models

(ends 4:00 PM)

Tutorial:

Jailbreaking LLMs and Agentic Systems: Attacks, Defenses, and Evaluations

(ends 4:00 PM)

Tutorial:

Flowing Through Continuous-Time Generative Models: A Clear and Systematic Tour

(ends 4:00 PM)

Tutorial:

Training Neural Networks at Any Scale

(ends 4:00 PM)

Tutorial:

The Underlying Logic of Language Models

(ends 4:00 PM)

4 p.m.

Break:

Coffee Break

(ends 5:00 PM)

Exhibits

(ends 8:00 PM)

Expo Demonstration:

IBM software agent for code (ISAC)

(ends 8:00 PM)

Expo Demonstration:

Production Strength Sales Agent

(ends 8:00 PM)

Expo Demonstration:

Excitement-Driven AI Sports Commentary Generation

(ends 8:00 PM)

Expo Demonstration:

Building Advanced Software Engineering LLM Agents with Codellm-Devkit (CLDK)

(ends 8:00 PM)

Expo Demonstration:

WCA4Z Platform - Accelerating Legacy Code Modernization Using AI Agents, Deep Program Analysis, and Scalable AI Pipelines

(ends 8:00 PM)

Expo Demonstration:

Prompt Declaration Language (PDL)

(ends 8:00 PM)

Expo Demonstration:

Real-World Autonomy: Building Modular, Voice-Guided Embodied Agents with SLMs and Vision

(ends 7:00 PM)

Expo Demonstration:

Knobs of the Mind: Dopamine, Serotonin, and a Maze-Running Rover

(ends 8:00 PM)

4:30 p.m.

Expo Workshop:

AI in Finance: Innovation & Emerging Opportunities

(ends 6:00 PM)

Expo Workshop:

Uncertainty Estimation in LLM-Generated Content

(ends 6:00 PM)

Expo Workshop:

Evaluation of GenAI models

(ends 6:00 PM)

Expo Talk Panel:

Building Production Ready Agentic Systems: Architecture, LLM-based Evaluation, and GRPO Training

(ends 5:30 PM)

Expo Talk Panel:

Graph Foundation Models: Thoughts and Results

(ends 6:00 PM)

6:30 p.m.

Reception:

Opening Remarks and Welcome Reception

(ends 8:00 PM)

TUE 15 JUL

7:30 a.m.

Registration West

(ends 6:00 PM)

Registration East

(ends 12:00 PM)

ICML Lounge Area:

ICML Lounge Area

(ends 7:00 PM)

8 a.m.

Affinity Workshop:

New In ML

(ends 5:30 PM)

8:30 a.m.

Invited Talk:

AI's Models of the World, and Ours

Jon Kleinberg

(ends 9:30 AM)

9:30 a.m.

Break:

Coffee Break

(ends 10:00 AM)

Exhibits

(ends 6:00 PM)

10 a.m.

Oral 1A Alignment and Agents [10:00-11:00]

Orals 10:00-11:00

[10:00] Multi-agent Architecture Search via Agentic Supernet

[10:15] Training a Generally Curious Agent

[10:30] Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

[10:45] CollabLLM: From Passive Responders to Active Collaborators

(ends 11:00 AM)

Oral 1B Positions: Better Ways to Do Machine Learning [10:00-11:00]

Orals 10:00-11:00

[10:00] Position: The AI Conference Peer Review Crisis Demands Author Feedback and Reviewer Rewards

[10:15] Position: Not All Explanations for Deep Learning Phenomena Are Equally Valuable

[10:30] Position: Certified Robustness Does Not (Yet) Imply Model Security

[10:45] Position: Probabilistic Modelling is Sufficient for Causal Inference

(ends 11:00 AM)

Oral 1C Applications in Computer Vision [10:00-11:00]

Orals 10:00-11:00

[10:00] VideoRoPE: What Makes for Good Video Rotary Position Embedding?

[10:15] ReferSplat: Referring Segmentation in 3D Gaussian Splatting

[10:30] Orthogonal Subspace Decomposition for Generalizable AI-Generated Image Detection

[10:45] VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models

(ends 11:00 AM)

Oral 1D Learning Dynamics 1 [10:00-11:00]

Orals 10:00-11:00

[10:00] Algorithm Development in Neural Networks: Insights from the Streaming Parity Task

[10:15] Learning Dynamics in Continual Pre-Training for Large Language Models

[10:30] Strategy Coopetition Explains the Emergence and Transience of In-Context Learning

[10:45] Transformative or Conservative? Conservation laws for ResNets and Transformers

(ends 11:00 AM)

Oral 1E Theory and Phenomenology [10:00-11:00]

Orals 10:00-11:00

[10:00] An analytic theory of creativity in convolutional diffusion models

[10:15] Layer by Layer: Uncovering Hidden Representations in Language Models

[10:30] Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks

[10:45] Emergence in non-neural models: grokking modular arithmetic via average gradient outer product

(ends 11:00 AM)

11 a.m.

Poster Session 1 West [11:00-1:30]

Posters 11:00-1:30

Autoformulation of Mathematical Optimization Models Using LLMs

Ladder-Residual: Parallelism-Aware Architecture for Accelerating Large Model Inference with Communication Overlapping

Delta Decompression for MoE-based LLMs Compression

Can Compressed LLMs Truly Act? An Empirical Evaluation of Agentic Capabilities in LLM Compression

MoE-SVD: Structured Mixture-of-Experts LLMs Compression via Singular Value Decomposition

When Every Millisecond Counts: Real-Time Anomaly Detection via the Multimodal Asynchronous Hybrid Network

Compressed and distributed least-squares regression: convergence rates with applications to federated learning

AKORN: Adaptive Knots generated Online for RegressioN splines

Pareto-Optimality, Smoothness, and Stochasticity in Learning-Augmented One-Max-Search

GaussMark: A Practical Approach for Structural Watermarking of Language Models

Identifying Metric Structures of Deep Latent Variable Models

TAROT: Targeted Data Selection via Optimal Transport

Core Knowledge Deficits in Multi-Modal Language Models

A Sharper Global Convergence Analysis for Average Reward Reinforcement Learning via an Actor-Critic Approach

The Sample Complexity of Online Strategic Decision Making with Information Asymmetry and Knowledge Transportability

Overcoming the Curse of Dimensionality in Reinforcement Learning Through Approximate Factorization

Can RLHF be More Efficient with Imperfect Reward Models? A Policy Coverage Perspective

Diffusion models for Gaussian distributions: Exact solutions and Wasserstein errors

Average Sensitivity of Hierarchical $k$-Median Clustering

Learning multivariate Gaussians with imperfect advice

On The Concurrence of Layer-wise Preconditioning Methods and Provable Feature Learning

Sparse-pivot: Dynamic correlation clustering for node insertions

Calibrating Video Watch-time Predictions with Credible Prototype Alignment

Incremental Gradient Descent with Small Epoch Counts is Surprisingly Slow on Ill-Conditioned Problems

Adaptive Self-improvement LLM Agentic System for ML Library Development

PEINR: A Physics-enhanced Implicit Neural Representation for High-Fidelity Flow Field Reconstruction

PDE-Transformer: Efficient and Versatile Transformers for Physics Simulations

Topology-aware Neural Flux Prediction Guided by Physics

OneForecast: A Universal Framework for Global and Regional Weather Forecasting

Efficient and Scalable Density Functional Theory Hamiltonian Prediction through Adaptive Sparsity

WyckoffDiff -- A Generative Diffusion Model for Crystal Symmetry

Zero-Shot Cyclic Peptide Design via Composable Geometric Constraints

Annealing Flow Generative Models Towards Sampling High-Dimensional and Multi-Modal Distributions

Sounding that Object: Interactive Object-Aware Image to Audio Generation

PINNsAgent: Automated PDE Surrogation with Large Language Models

Temperature-Annealed Boltzmann Generators

Arbitrarily-Conditioned Multi-Functional Diffusion for Multi-Physics Emulation

A Machine Learning Approach to Duality in Statistical Physics

Tensor-Var: Efficient Four-Dimensional Variational Data Assimilation

Multi-Timescale Dynamics Model Bayesian Optimization for Plasma Stabilization in Tokamaks

MIPT: Multilevel Informed Prompt Tuning for Robust Molecular Property Prediction

Open Materials Generation with Stochastic Interpolants

On Explaining Equivariant Graph Networks via Improved Relevance Propagation

Relational Invariant Learning for Robust Solvation Free Energy Prediction

Interpolating Neural Network-Tensor Decomposition (INN-TD): a scalable and interpretable approach for large-scale physics-based problems

HybridGS: High-Efficiency Gaussian Splatting Data Compression using Dual-Channel Sparse Representation and Point Cloud Encoder

CostFilter-AD: Enhancing Anomaly Detection through Matching Cost Filtering

HetSSNet: Spatial-Spectral Heterogeneous Graph Learning Network for Panchromatic and Multispectral Images Fusion

MIRROR: Make Your Object-Level Multi-View Generation More Consistent with Training-Free Rectification

Generative Point Cloud Registration

Feature out! Let Raw Image as Your Condition for Blind Face Restoration

VTGaussian-SLAM: RGBD SLAM for Large Scale Scenes with Splatting View-Tied 3D Gaussians

VideoRoPE: What Makes for Good Video Rotary Position Embedding?

ReferSplat: Referring Segmentation in 3D Gaussian Splatting

Privacy-Shielded Image Compression: Defending Against Exploitation from Vision-Language Pretrained Models

Cavia: Camera-controllable Multi-view Video Diffusion with View-Integrated Attention

Asymmetric Decision-Making in Online Knowledge Distillation: Unifying Consensus and Divergence

Long-Term TalkingFace Generation via Motion-Prior Conditional Diffusion Model

Efficient Motion Prompt Learning for Robust Visual Tracking

PhySpec: Physically Consistent Spectral Reconstruction via Orthogonal Subspace Decomposition and Self-Supervised Meta-Auxiliary Learning

PISA Experiments: Exploring Physics Post-Training for Video Diffusion Models by Watching Stuff Drop

SafeMap: Robust HD Map Construction from Incomplete Observations

When Model Knowledge meets Diffusion Model: Diffusion-assisted Data-free Image Synthesis with Alignment of Domain and Class

Learning Event Completeness for Weakly Supervised Video Anomaly Detection

Visual Autoregressive Modeling for Image Super-Resolution

Beyond Confidence: Exploiting Homogeneous Pattern for Semi-Supervised Semantic Segmentation

UniMC: Taming Diffusion Transformer for Unified Keypoint-Guided Multi-Class Image Generation

Orthogonal Subspace Decomposition for Generalizable AI-Generated Image Detection

Unifying 2D and 3D Vision-Language Understanding

DyPolySeg: Taylor Series-Inspired Dynamic Polynomial Fitting Network for Few-shot Point Cloud Semantic Segmentation

sciLaMA: A Single-Cell Representation Learning Framework to Leverage Prior Knowledge from Large Language Models

CellFlux: Simulating Cellular Morphology Changes via Flow Matching

Raptor: Scalable Train-Free Embeddings for 3D Medical Volumes Leveraging Pretrained 2D Foundation Models

All-atom inverse protein folding through discrete flow matching

Aligning Protein Conformation Ensemble Generation with Physical Feedback

CombiMOTS: Combinatorial Multi-Objective Tree Search for Dual-Target Molecule Generation

MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding

UniMoMo: Unified Generative Modeling of 3D Molecules for De Novo Binder Design

Identifying biological perturbation targets through causal differential networks

Learning Invariant Causal Mechanism from Vision-Language Models

BounDr.E: Predicting Drug-likeness via Biomedical Knowledge Alignment and EM-like One-Class Boundary Optimization

Multi-Marginal Stochastic Flow Matching for High-Dimensional Snapshot Data at Irregular Time Points

InfoSEM: A Deep Generative Model with Informative Priors for Gene Regulatory Network Inference

A Variational Perspective on Generative Protein Fitness Optimization

Do Multiple Instance Learning Models Transfer?

Reliable Algorithm Selection for Machine Learning-Guided Design

"Why Is There a Tumor?": Tell Me the Reason, Show Me the Evidence

ViTally Consistent: Scaling Biological Representation Learning for Cell Microscopy

Visual and Domain Knowledge for Professional-level Graph-of-Thought Medical Reasoning

A Cross Modal Knowledge Distillation & Data Augmentation Recipe for Improving Transcriptomics Representations through Morphological Features

Unified Screening for Multiple Diseases

BinauralFlow: A Causal and Streamable Approach for High-Quality Binaural Speech Synthesis with Flow Matching Models

UI-Vision: A Desktop-centric GUI Benchmark for Visual Perception and Interaction

Context is Key: A Benchmark for Forecasting with Essential Textual Information

CMoS: Rethinking Time Series Prediction Through the Lens of Chunk-wise Spatial Correlations

TimePoint: Accelerated Time Series Alignment via Self-Supervised Keypoint and Descriptor Learning

LOB-Bench: Benchmarking Generative AI for Finance - an Application to Limit Order Book Data

ProDiff: Prototype-Guided Diffusion for Minimal Information Trajectory Imputation

Efficient Personalized Adaptation for Physiological Signal Foundation Model

Towards Learning to Complete Anything in Lidar

Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations

OTTER: A Vision-Language-Action Model with Text-Aware Visual Feature Extraction

SAH-Drive: A Scenario-Aware Hybrid Planner for Closed-Loop Vehicle Trajectory Generation

DexScale: Automating Data Scaling for Sim2Real Generalizable Robot Control

Self-cross Feature based Spiking Neural Networks for Efficient Few-shot Learning

Testing the Limits of Fine-Tuning for Improving Visual Cognition in Vision Language Models

Pre-Training Graph Contrastive Masked Autoencoders are Strong Distillers for EEG

Faster and Stronger: When ANN-SNN Conversion Meets Parallel Spiking Calculation

Fleet of Agents: Coordinated Problem Solving with Large Language Models

A Reasoning-Based Approach to Cryptic Crossword Clue Solving

MoHAVE: Mixture of Hierarchical Audio-Visual Experts for Robust Speech Recognition

Policy Filtration for RLHF to Mitigate Noise in Reward Models

LaRA: Benchmarking Retrieval-Augmented Generation and Long-Context LLMs – No Silver Bullet for LC or RAG Routing

Feature-Mapping Topology Optimization with Neural Heaviside Signed Distance Functions

Universal Biological Sequence Reranking for Improved De Novo Peptide Sequencing

Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction

LaMAGIC2: Advanced Circuit Formulations for Language Model-Based Analog Topology Generation

Geometry-Informed Neural Networks

Flat-LoRA: Low-Rank Adaptation over a Flat Loss Landscape

Efficient Distributed Optimization under Heavy-Tailed Noise

Online Conformal Prediction via Online Optimization

Learning to Generate Projections for Reducing Dimensionality of Heterogeneous Linear Programming Problems

Schwarz–Schur Involution: Lightspeed Differentiable Sparse Linear Solvers

Decoupled SGDA for Games with Intermittent Strategy Communication

Momentum-Driven Adaptivity: Towards Tuning-Free Asynchronous Federated Learning

Lean and Mean Adaptive Optimization via Subset-Norm and Subspace-Momentum with Convergence Guarantees

Efficient Curvature-Aware Hypergradient Approximation for Bilevel Optimization

Automatic Differentiation of Optimization Algorithms with Time-Varying Updates

Secant Line Search for Frank-Wolfe Algorithms

Active Learning of Deep Neural Networks via Gradient-Free Cutting Planes

Scalable Approximation Algorithms for $p$-Wasserstein Distance and Its Variants

SHIELD: Multi-task Multi-distribution Vehicle Routing Solver with Sparsity and Hierarchy

An Asymptotically Optimal Approximation Algorithm for Multiobjective Submodular Maximization at Scale

Discrepancy Minimization in Input-Sparsity Time

Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces

Scaling Value Iteration Networks to 5000 Layers for Extreme Long-Term Planning

Action-Constrained Imitation Learning

A Reduction Framework for Distributionally Robust Reinforcement Learning under Average Reward

Off-Policy Actor-Critic for Adversarial Observation Robustness: Virtual Alternative Training via Symmetric Policy Evaluation

Adaptive Exploration for Multi-Reward Multi-Policy Evaluation

A Sub-Problem Quantum Alternating Operator Ansatz for Correlation Clustering

Towards Universal Offline Black-Box Optimization via Learning Language Model Embeddings

Improved Lower Bounds for First-order Stochastic Non-convex Optimization under Markov Sampling

A Memory Efficient Randomized Subspace Optimization Method for Training Large Language Models

Convergence of Mean-Field Langevin Stochastic Descent-Ascent for Distributional Minimax Optimization

Optimal Transport Barycenter via Nonconvex-Concave Minimax Optimization

A Novel Characterization of the Population Area Under the Risk Coverage Curve (AURC) and Rates of Finite Sample Estimators

MetaOptimize: A Framework for Optimizing Step Sizes and Other Meta-parameters

Constant Stepsize Local GD for Logistic Regression: Acceleration by Instability

Distributed Retraction-Free and Communication-Efficient Optimization on the Stiefel Manifold

Ringmaster ASGD: The First Asynchronous SGD with Optimal Time Complexity

FedOne: Query-Efficient Federated Learning for Black-box Discrete Prompt Learning

PipeOffload: Improving Scalability of Pipeline Parallelism with Memory Optimization

FRUGAL: Memory-Efficient Optimization by Reducing State Overhead for Scalable Training

Riemannian Diffusion Adaptation for Distributed Optimization on Manifolds

Differentiable Quadratic Optimization For the Maximum Independent Set Problem

QPRL : Learning Optimal Policies with Quasi-Potential Functions for Asymmetric Traversal

In-Context Reinforcement Learning From Suboptimal Historical Data

Calibrated Value-Aware Model Learning with Probabilistic Environment Models

Time-Aware World Model for Adaptive Prediction and Control

Pessimism Principle Can Be Effective: Towards a Framework for Zero-Shot Transfer Reinforcement Learning

Log-Sum-Exponential Estimator for Off-Policy Evaluation and Learning

Temporal Distance-aware Transition Augmentation for Offline Model-based Reinforcement Learning

Reflect-then-Plan: Offline Model-Based Planning through a Doubly Bayesian Lens

Multi-Stage Manipulation with Demonstration-Augmented Reward, Policy, and World Model Learning

Trajectory World Models for Heterogeneous Environments

Stealing That Free Lunch: Exposing the Limits of Dyna-Style Reinforcement Learning

The Impact of On-Policy Parallelized Data Collection on Deep Reinforcement Learning Networks

Robust Reward Alignment via Hypothesis Space Batch Cutting

Diversifying Robot Locomotion Behaviors with Extrinsic Behavioral Curiosity

Agent-Centric Actor-Critic for Asynchronous Multi-Agent Reinforcement Learning

Ad-Hoc Human-AI Coordination Challenge

Revisiting Cooperative Off-Policy Multi-Agent Reinforcement Learning

R3DM: Enabling Role Discovery and Diversity Through Dynamics Models in Multi-agent Reinforcement Learning

AssistanceZero: Scalably Solving Assistance Games

Goal-Space Planning with Subgoal Models

Faster Approximation Algorithms for k-Center via Data Reduction

Private Lossless Multiple Release

Batch List-Decodable Linear Regression via Higher Moments

Model Uncertainty Quantification by Conformal Prediction in Continual Learning

Understanding the Forgetting of (Replay-based) Continual Learning via Feature Learning: Angle Matters

Uniform Mean Estimation for Heavy-Tailed Distributions via Median-of-Means

On the Statistical Mechanisms of Distributional Compositional Generalization

Securing Equal Share: A Principled Approach for Learning Multiplayer Symmetric Games

Self-Play $Q$-Learners Can Provably Collude in the Iterated Prisoner's Dilemma

Fraud-Proof Revenue Division on Subscription Platforms

The impact of uncertainty on regularized learning in games

Solving Zero-Sum Convex Markov Games

Preference-CFR: Beyond Nash Equilibrium for Better Game Strategies

Settling the Maximin Share Fairness for Scheduling among Groups of Machines

Contract Design Under Approximate Best Responses

COSDA: Counterfactual-based Susceptibility Risk Framework for Open-Set Domain Adaptation

The Role of Sparsity for Length Generalization in LLMs

Optimal Transfer Learning for Missing Not-at-Random Matrix Completion

When Can Proxies Improve the Sample Complexity of Preference Learning?

Representations Shape Weak-to-Strong Generalization: Theoretical Insights and Empirical Predictions

Time to Spike? Understanding the Representational Power of Spiking Neural Networks in Discrete Time

Beyond Bradley-Terry Models: A General Preference Model for Language Model Alignment

Provable In-Context Vector Arithmetic via Retrieving Task Concepts

Algorithm Development in Neural Networks: Insights from the Streaming Parity Task

The Role of Randomness in Stability

Approximation to Smooth Functions by Low-Rank Swish Networks

On the Learnability of Distribution Classes with Adaptive Adversaries

On the Training Convergence of Transformers for In-Context Classification of Gaussian Mixtures

On Learning Parallel Pancakes with Mostly Uniform Weights

How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias

Representation Preserving Multiclass Agnostic to Realizable Reduction

Constrained Pareto Set Identification with Bandit Feedback

Multi-Armed Bandits with Interference: Bridging Causal Inference and Adversarial Bandits

A Trichotomy for List Transductive Online Learning

Offline Learning for Combinatorial Multi-armed Bandits

Nearly Optimal Algorithms for Contextual Dueling Bandits from Adversarial Feedback

Tracking Most Significant Shifts in Infinite-Armed Bandits

Online Learning in the Random-Order Model

Optimal and Practical Batched Linear Bandit Algorithm

Contextual Online Decision Making with Infinite-Dimensional Functional Regression

Optimal Algorithm for Max-Min Fair Bandit

The Batch Complexity of Bandit Pure Exploration

Adaptive Sample Sharing for Multi Agent Linear Bandits

Policy Optimization for CMDPs with Bandit Feedback: Learning Stochastic and Adversarial Constraints

Contextual Linear Bandits with Delay as Payoff

Leveraging Model Guidance to Extract Training Data from Personalized Diffusion Models

(ends 1:30 PM)

Poster Session 1 East [11:00-1:30]

Posters 11:00-1:30

CollabLLM: From Passive Responders to Active Collaborators

The Disparate Benefits of Deep Ensembles

Relative Error Fair Clustering in the Weak-Strong Oracle Model

FDGen: A Fairness-Aware Graph Generation Model

GHOST: Generalizable One-Shot Federated Graph Learning with Proxy-Based Topology Knowledge Retention

CAN: Leveraging Clients As Navigators for Generative Replay in Federated Continual Learning

Cape: Context-Aware Prompt Perturbation Mechanism with Differential Privacy

Privacy Amplification by Structured Subsampling for Deep Differentially Private Time Series Forecasting

Tracking The Best Expert Privately

Privacy Amplification Through Synthetic Data: Insights from Linear Regression

Layer-wise Alignment: Examining Safety Alignment Across Image Encoder Layers in Vision Language Models

Preference learning made easy: Everything should be understood through win rate

Vision-Language Models Create Cross-Modal Task Representations

Detecting Strategic Deception with Linear Probes

Which Attention Heads Matter for In-Context Learning?

Tensor Product Neural Networks for Functional ANOVA Model

From Pixels to Perception: Interpretable Predictions via Instance-wise Grouped Feature Selection

Fast Estimation of Partial Dependence Functions using Trees

AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoders

On the Interplay between Graph Structure and Learning Algorithms in Graph Neural Networks

Data-driven Design of Randomized Control Trials with Guaranteed Treatment Effects

SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability

Learning Multi-Level Features with Matryoshka Sparse Autoencoders

Towards Attributions of Input Variables in a Coalition

Auditing Prompt Caching in Language Model APIs

Towards Global-level Mechanistic Interpretability: A Perspective of Modular Circuits of Large Language Models

(How) Can Transformers Predict Pseudo-Random Numbers?

Prediction via Shapley Value Regression

Beyond Induction Heads: In-Context Meta Learning Induces Multi-Phase Circuit Emergence

Domain2Vec: Vectorizing Datasets to Find the Optimal Data Mixture without Training

On the Impact of Performative Risk Minimization for Binary Random Variables

A Mathematical Framework for AI-Human Integration in Work

Neural Guided Diffusion Bridges

Learning-Order Autoregressive Models with Application to Molecular Graph Generation

Revisiting Unbiased Implicit Variational Inference

Federated Generalised Variational Inference: A Robust Probabilistic Federated Learning Framework

Aggregation of Dependent Expert Distributions in Multimodal Variational Autoencoders

Exploring Large Action Sets with Hyperspherical Embeddings using von Mises-Fisher Sampling

A General Framework for Inference-time Scaling and Steering of Diffusion Models

Tuning Sequential Monte Carlo Samplers via Greedy Incremental Divergence Minimization

Learning Along the Arrow of Time: Hyperbolic Geometry for Backward-Compatible Representation Learning

Fast and Provable Algorithms for Sparse PCA with Improved Sample Complexity

Wrapped Gaussian on the manifold of Symmetric Positive Definite Matrices

Doubly Robust Conformalized Survival Analysis with Right-Censored Data

Do Bayesian Neural Networks Actually Behave Like Bayesian Models?

BILBO: BILevel Bayesian Optimization

Integration-free Kernels for Equivariant Gaussian Process Modelling

Robust and Conjugate Spatio-Temporal Gaussian Processes

Pareto-frontier Entropy Search with Variational Lower Bound Maximization

Revisiting Non-Acyclic GFlowNets in Discrete Environments

When to retrain a machine learning model

Poly2Vec: Polymorphic Fourier-Based Encoding of Geospatial Objects for GeoAI Applications

LBI-FL: Low-Bit Integerized Federated Learning with Temporally Dynamic Bit-Width Allocation

Navigating Conflicting Views: Harnessing Trust for Learning

Nonparametric Teaching for Graph Property Learners

Confidence Difference Reflects Various Supervised Signals in Confidence-Difference Classification

Learning to Match Unpaired Data with Minimum Entropy Coupling

Universal Neural Optimal Transport

Random Registers for Cross-Domain Few-Shot Learning

Test-Time Selective Adaptation for Uni-Modal Distribution Shift in Multi-Modal Data

Instance Correlation Graph-based Naive Bayes

Ab Initio Nonparametric Variable Selection for Scalable Symbolic Regression with Large $p$

Knowledge-Guided Wasserstein Distributionally Robust Optimization

Does learning the right latent variables necessarily improve in-context learning?

Update Your Transformer to the Latest Release: Re-Basin of Task Vectors

R2-T2: Re-Routing in Test-Time for Multimodal Mixture-of-Experts

iDPA: Instance Decoupled Prompt Attention for Incremental Medical Object Detection

Be Confident: Uncovering Overfitting in MLLM Multi-Task Tuning

Homophily Enhanced Graph Domain Adaptation

Revisiting Continuity of Image Tokens for Cross-domain Few-shot Learning

Divide and Conquer: Learning Label Distribution with Subtasks

GRAIL: Graph Edit Distance and Node Alignment using LLM-Generated Code

Soup-of-Experts: Pretraining Specialist Models via Parameters Averaging

Randomized Dimensionality Reduction for Euclidean Maximization and Diversity Measures

Learning Optimal Multimodal Information Bottleneck Representations

Diss-l-ECT: Dissecting Graph Data with Local Euler Characteristic Transforms

Latent Score-Based Reweighting for Robust Classification on Imbalanced Tabular Data

Deep Unsupervised Hashing via External Guidance

GMAIL: Generative Modality Alignment for generated Image Learning

Aligning Multimodal Representations through an Information Bottleneck

Uncertainty Quantification for LLM-Based Survey Simulations

CogMath: Assessing LLMs' Authentic Mathematical Ability from a Human Cognitive Perspective

Measuring Diversity in Synthetic Datasets

Copilot Arena: A Platform for Code LLM Evaluation in the Wild

LMAct: A Benchmark for In-Context Imitation Learning with Long Multimodal Demonstrations

Toward Efficient Kernel-Based Solvers for Nonlinear PDEs

Statistical and Computational Guarantees of Kernel Max-Sliced Wasserstein Distances

Competing Bandits in Matching Markets via Super Stability

Comparing Few to Rank Many: Active Human Preference Learning Using Randomized Frank-Wolfe Method

Non-stationary Online Learning for Curved Losses: Improved Dynamic Regret via Mixability

Active feature acquisition via explainability-driven ranking

Clustering Items through Bandit Feedback: Finding the Right Feature out of Many

Preconditioned Riemannian Gradient Descent Algorithm for Low-Multilinear-Rank Tensor Completion

Deep Streaming View Clustering

Signed Laplacians for Constrained Graph Clustering

Revisiting Instance-Optimal Cluster Recovery in the Labeled Stochastic Block Model

Arrow: Accelerator for Time Series Causal Discovery with Time Weaving

Quantifying Treatment Effects: Estimating Risk Ratios via Observational Studies

A Recipe for Causal Graph Regression: Confounding Effects Revisited

Isolated Causal Effects of Natural Language

Distinguishing Cause from Effect with Causal Velocity Models

Data-Driven Selection of Instrumental Variables for Additive Nonlinear, Constant Effects Models

Counterfactual Effect Decomposition in Multi-Agent Sequential Decision Making

A Causal World Model Underlying Next Token Prediction: Exploring GPT in a Controlled Environment

Telling Peer Direct Effects from Indirect Effects in Observational Network Data

A Sample Efficient Conditional Independence Test in the Presence of Discretization

BoA: Attention-aware Post-training Quantization without Backpropagation

Tackling Dimensional Collapse toward Comprehensive Universal Domain Adaptation

Learning without Isolation: Pathway Protection for Continual Learning

DOLPHIN: A Programmable Framework for Scalable Neurosymbolic Learning

Self-Improving Transformers Overcome Easy-to-Hard and Length Generalization Challenges

A Closer Look at Generalized BH Algorithm for Out-of-Distribution Detection

Enhancing Adversarial Robustness with Conformal Prediction: A Framework for Guaranteed Model Reliability

Bifurcate then Alienate: Incomplete Multi-view Clustering via Coupled Distribution Learning with Linear Overhead

FedSSI: Rehearsal-Free Continual Federated Learning with Synergistic Synaptic Intelligence

Joint Metric Space Embedding by Unbalanced Optimal Transport with Gromov–Wasserstein Marginal Penalization

Improving the Variance of Differentially Private Randomized Experiments through Clustering

Doubly Protected Estimation for Survival Outcomes Utilizing External Controls for Randomized Clinical Trials

Balancing Interference and Correlation in Spatial Experimental Designs: A Causal Graph Cut Approach

AtlasD: Automatic Local Symmetry Discovery

Investigating the Overlooked Hessian Structure: From CNNs to LLMs

Concept-Centric Token Interpretation for Vector-Quantized Generative Models

CtrlSynth: Controllable Image Text Synthesis for Data-Efficient Multimodal Learning

A Simple Model of Inference Scaling Laws

Minimum Width for Universal Approximation using Squashable Activation Functions

Sparse Training from Random Initialization: Aligning Lottery Ticket Masks using Weight Symmetry

On the Power of Context-Enhanced Learning in LLMs

Beyond Zero Initialization: Investigating the Impact of Non-Zero Initialization on LoRA Fine-Tuning Dynamics

When Diffusion Models Memorize: Inductive Biases in Probability Flow of Minimum-Norm Shallow Neural Nets

Neural Collapse Beyond the Unconstrained Features Model: Landscape, Dynamics, and Generalization in the Mean-Field Regime

Transformative or Conservative? Conservation laws for ResNets and Transformers

Global curvature for second-order optimization of neural networks

TabFSBench: Tabular Benchmark for Feature Shifts in Open Environments

Advancing Personalized Learning with Neural Collapse for Long-Tail Challenge

Towards Robust Influence Functions with Flat Validation Minima

Test-Time Canonicalization by Foundation Models for Robust Perception

Bi-perspective Splitting Defense: Achieving Clean-Seed-Free Backdoor Security

LAION-C: An Out-of-Distribution Benchmark for Web-Scale Vision Models

Adversarial Perturbations Are Formed by Iteratively Learning Linear Combinations of the Right Singular Vectors of the Adversarial Jacobian

OOD-Chameleon: Is Algorithm Selection for OOD Generalization Learnable?

MTSTRec: Multimodal Time-Aligned Shared Token Recommender

Lightweight Online Adaption for Time Series Foundation Model Forecasts

$K^2$VAE: A Koopman-Kalman Enhanced Variational AutoEncoder for Probabilistic Time Series Forecasting

Winner-takes-all for Multivariate Probabilistic Time Series Forecasting

An analytic theory of creativity in convolutional diffusion models

Optimizing Robustness and Accuracy in Mixture of Experts: A Dual-Model Approach

Quadratic Upper Bound for Boosting Robustness

From Debate to Equilibrium: Belief‑Driven Multi‑Agent LLM Reasoning via Bayesian Nash Equilibrium

An Adaptive Orthogonal Convolution Scheme for Efficient and Flexible CNN Architectures

Boosting Adversarial Robustness with CLAT: Criticality Leveraged Adversarial Training

SGD Jittering: A Training Strategy for Robust and Accurate Model-Based Architectures

Stochastic Layer-Wise Shuffle for Improving Vision Mamba Training

Mechanistic PDE Networks for Discovery of Governing Equations

The Logical Implication Steering Method for Conditional Interventions on Transformer Generation

PRIME: Deep Imbalanced Regression with Proxies

Deep Principal Support Vector Machines for Nonlinear Sufficient Dimension Reduction

Double-Filter: Efficient Fine-tuning of Pre-trained Vision-Language Models via Patch&Layer Filtering

Emergence and Effectiveness of Task Vectors in In-Context Learning: An Encoder Decoder Perspective

Large Language Models to Diffusion Finetuning

Peripheral Memory for LLMs: Integration of Sequential Memory Banks with Adaptive Querying

The Synergy of LLMs & RL Unlocks Offline Learning of Generalizable Language-Conditioned Policies with Low-fidelity Data

Prune 'n Predict: Optimizing LLM Decision-making with Conformal Prediction

Pivoting Factorization: A Compact Meta Low-Rank Representation of Sparsity for Efficient Inference in Large Language Models

OrthoRank: Token Selection via Sink Token Orthogonality for Efficient LLM inference

Restoring Calibration for Aligned Large Language Models: A Calibration-Aware Fine-Tuning Approach

Soft Reasoning: Navigating Solution Spaces in Large Language Models through Controlled Embedding Exploration

The Missing Alignment Link of In-context Learning on Sequences

RE-Bench: Evaluating Frontier AI R&D Capabilities of Language Model Agents against Human Experts

Improving Parallel Program Performance with LLM Optimizers via Agent-System Interfaces

Great Models Think Alike and this Undermines AI Oversight

Tool Unlearning for Tool-Augmented LLMs

EffiCoder: Enhancing Code Generation in Large Language Models through Efficiency-Aware Fine-tuning

RLTHF: Targeted Human Feedback for LLM Alignment

PhantomWiki: On-Demand Datasets for Reasoning and Retrieval Evaluation

Catch Your Emotion: Sharpening Emotion Perception in Multimodal Large Language Models

LLM Alignment as Retriever Optimization: An Information Retrieval Perspective

A Hitchhiker's Guide to Scaling Law Estimation

From RAG to Memory: Non-Parametric Continual Learning for Large Language Models

SepLLM: Accelerate Large Language Models by Compressing One Segment into One Separator

BRiTE: Bootstrapping Reinforced Thinking Process to Enhance Language Model Reasoning

AMPO: Active Multi Preference Optimization for Self-play Preference Selection

How Much Can We Forget about Data Contamination?

RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning

Learning to Keep a Promise: Scaling Language Model Decoding Parallelism with Learned Asynchronous Decoding

Fourier Position Embedding: Enhancing Attention’s Periodic Extension for Length Generalization

GuidedQuant: Large Language Model Quantization via Exploiting End Loss Guidance

Large Language Models are Demonstration Pre-Selectors for Themselves

LIFT the Veil for the Truth: Principal Weights Emerge after Rank Reduction for Reasoning-Focused Supervised Fine-Tuning

Speculate, then Collaborate: Fusing Knowledge of Language Models during Decoding

Diving into Self-Evolving Training for Multimodal Reasoning

Layer by Layer: Uncovering Hidden Representations in Language Models

Learning Dynamics in Continual Pre-Training for Large Language Models

Training a Generally Curious Agent

TGDPO: Harnessing Token-Level Reward Guidance for Enhancing Direct Preference Optimization

Residual Matrix Transformers: Scaling the Size of the Residual Stream

PILAF: Optimal Human Preference Sampling for Reward Modeling

Quamba2: A Robust and Scalable Post-training Quantization Framework for Selective State Space Models

Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

The Surprising Effectiveness of Test-Time Training for Few-Shot Learning

From Passive to Active Reasoning: Can Large Language Models Ask the Right Questions under Incomplete Information?

Function-to-Style Guidance of LLMs for Code Translation

ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verification

Self-Disentanglement and Re-Composition for Cross-Domain Few-Shot Segmentation

The Canary’s Echo: Auditing Privacy Risks of LLM-Generated Synthetic Text

MATS: An Audio Language Model under Text-only Supervision

HPS: Hard Preference Sampling for Human Preference Alignment

Understanding the Skill Gap in Recurrent Language Models: The Role of the Gather-and-Aggregate Mechanism

Focus On This, Not That! Steering LLMs with Adaptive Feature Specification

SelfCite: Self-Supervised Alignment for Context Attribution in Large Language Models

On-the-Fly Adaptive Distillation of Transformer to Dual-State Linear Attention for Long-Context LLM Serving

Reasoning Through Execution: Unifying Process and Outcome Rewards for Code Generation

Adapting While Learning: Grounding LLMs for Scientific Problems with Tool Usage Adaptation

ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference

Adapter Naturally Serves as Decoupler for Cross-Domain Few-Shot Semantic Segmentation

Splitting with Importance-aware Updating for Heterogeneous Federated Learning with Large Language Models

Bring Reason to Vision: Understanding Perception and Reasoning through Model Merging

Towards Cost-Effective Reward Guided Text Generation

Synthetic Text Generation for Training Large Language Models via Gradient Matching

Nemotron-CORTEXA: Enhancing LLM Agents for Software Engineering Tasks via Improved Localization and Solution Diversity

Multi-agent Architecture Search via Agentic Supernet

From Language Models over Tokens to Language Models over Characters

Adaptive Localization of Knowledge Negation for Continual LLM Unlearning

RuleAdapter: Dynamic Rules for training Safety Reward Models in RLHF

FlatQuant: Flatness Matters for LLM Quantization

Mitigating Heterogeneous Token Overfitting in LLM Knowledge Editing

Rethinking Addressing in Language Models via Contextualized Equivariant Positional Encoding

Scaling Inference-Efficient Language Models

EvFocus: Learning to Reconstruct Sharp Images from Out-of-Focus Event Streams

A Stronger Mixture of Low-Rank Experts for Fine-Tuning Foundation Models

KVTuner: Sensitivity-Aware Layer-Wise Mixed-Precision KV Cache Quantization for Efficient and Nearly Lossless LLM Inference

PARM: Multi-Objective Test-Time Alignment via Preference-Aware Autoregressive Reward Model

Analytical Construction on Geometric Architectures: Transitioning from Static to Temporal Link Prediction

A Dynamical Systems-Inspired Pruning Strategy for Addressing Oversmoothing in Graph Attention Networks

G-Adaptivity: optimised graph-based mesh relocation for finite element methods

FicGCN: Unveiling the Homomorphic Encryption Efficiency from Irregular Graph Convolutional Networks

Rhomboid Tiling for Geometric Graph Deep Learning

Primal-Dual Neural Algorithmic Reasoning

A Cognac Shot To Forget Bad Memories: Corrective Unlearning for Graph Neural Networks

Balancing Efficiency and Expressiveness: Subgraph GNNs with Walk-Based Centrality

No Metric to Rule Them All: Toward Principled Evaluations of Graph-Learning Datasets

Hyperbolic-PDE GNN: Spectral Graph Neural Networks in the Perspective of A System of Hyperbolic Partial Differential Equations

Learning Latent Graph Structures and their Uncertainty

Propagate and Inject: Revisiting Propagation-Based Feature Imputation for Graphs with Partially Observed Features

Unifews: You Need Fewer Operations for Efficient Graph Neural Networks

iN2V: Bringing Transductive Node Embeddings to Inductive Graphs

TMetaNet: Topological Meta-Learning Framework for Dynamic Link Prediction

Aggregation Buffer: Revisiting DropEdge with a New Parameter Block

Simple Path Structural Encoding for Graph Transformers

Less is More: Federated Graph Learning with Alleviating Topology Heterogeneity from A Causal Perspective

Towards a Unified Framework of Clustering-based Anomaly Detection

Diffuse Everything: Multimodal Diffusion Models on Arbitrary State Spaces

VCT: Training Consistency Models with Variational Noise Coupling

Angle Domain Guidance: Latent Diffusion Requires Rotation Rather Than Extrapolation

Upcycling Text-to-Image Diffusion Models for Multi-Task Capabilities

Spherical-Nested Diffusion Model for Panoramic Image Outpainting

FlowAR: Scale-wise Autoregressive Image Generation Meets Flow Matching

Inference-Time Alignment of Diffusion Models with Direct Noise Optimization

Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator

Multidimensional Adaptive Coefficient for Inference Trajectory Optimization in Flow and Diffusion

VideoJAM: Joint Appearance-Motion Representations for Enhanced Motion Generation in Video Models

FlexiClip: Locality-Preserving Free-Form Character Animation

Sparse Autoencoders, Again?

Ensemble Distribution Distillation via Flow Matching

RIFLEx: A Free Lunch for Length Extrapolation in Video Diffusion Transformers

Generalized Interpolating Discrete Diffusion

How to Move Your Dragon: Text-to-Motion Synthesis for Large-Vocabulary Objects

Explainable Concept Generation through Vision-Language Preference Learning for Understanding Neural Networks' Internal Representations

Conditional Diffusion Model with Nonlinear Data Transformation for Time Series Forecasting

Masked Generative Nested Transformers with Decode Time Scaling

Efficient Generative Modeling with Residual Vector Quantization-Based Tokens

Masked Autoencoders Are Effective Tokenizers for Diffusion Models

Learning Monotonic Probabilities with a Generative Cost Model

Learning to Quantize for Training Vector-Quantized Networks

The Hidden Life of Tokens: Reducing Hallucination of Large Vision-Language Models Via Visual Information Steering

Expert Race: A Flexible Routing Strategy for Scaling Diffusion Transformer with Mixture of Experts

Learn from Downstream and Be Yourself in Multimodal Large Language Models Fine-Tuning

Towards Efficient Online Tuning of VLM Agents via Counterfactual Soft Reinforcement Learning

What If We Recaption Billions of Web Images with LLaMA-3?

BalancEdit: Dynamically Balancing the Generality-Locality Trade-off in Multi-modal Model Editing

Reasoning Limitations of Multimodal Large Language Models. A case study of Bongard Problems

Bongard in Wonderland: Visual Puzzles that Still Make AI Go Mad?

Griffin: Towards a Graph-Centric Relational Database Foundation Model

In-Context Learning as Conditioned Associative Memory Retrieval

Learning to (Learn at Test Time): RNNs with Expressive Hidden States

Efficient Parallel Training Methods for Spiking Neural Networks with Constant Time Complexity

LapSum - One Method to Differentiate Them All: Ranking, Sorting and Top-k Selection

Improving Memory Efficiency for Training KANs via Meta Learning

Eigenspectrum Analysis of Neural Networks without Aspect Ratio Bias

A First-order Generative Bilevel Optimization Framework for Diffusion Models

Training Deep Learning Models with Norm-Constrained LMOs

The underlying structures of self-attention: symmetry, directionality, and emergent dynamics in Transformer training

Strategy Coopetition Explains the Emergence and Transience of In-Context Learning

Selective Prompt Anchoring for Code Generation

On Expressive Power of Looped Transformers: Theoretical Analysis and Enhancement via Timestep Encoding

HashAttention: Semantic Sparsity for Faster Inference

Fast Inference with Kronecker-Sparse Matrices

KIND: Knowledge Integration and Diversion for Training Decomposable Models

Dynamic Sparse Training of Diagonally Sparse Networks

High Dynamic Range Novel View Synthesis with Single Exposure

Avoiding spurious sharpness minimization broadens applicability of SAM

On Exact Bit-level Reversible Transformers Without Changing Architecture

MixMin: Finding Data Mixtures via Convex Minimization

Rethinking the Stability-Plasticity Trade-off in Continual Learning from an Architectural Perspective

How Compositional Generalization and Creativity Improve as Diffusion Models are Trained

Towards Understanding Fine-Tuning Mechanisms of LLMs via Circuit Analysis

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks

How to set AdamW's weight decay as you scale model and dataset size

Erwin: A Tree-based Hierarchical Transformer for Large-scale Physical Systems

Stable Fair Graph Representation Learning with Lipschitz Constraint

TinyMIG: Transferring Generalization from Vision Foundation Models to Single-Domain Medical Imaging

Topological Signatures of Adversaries in Multimodal Alignments

The Panaceas for Improving Low-Rank Decomposition in Communication-Efficient Federated Learning

Kona: An Efficient Privacy-Preservation Framework for KNN Classification by Communication Optimization

Vision-Language Model Selection and Reuse for Downstream Adaptation

Universal Approximation of Mean-Field Models via Transformers

ZeroFlow: Overcoming Catastrophic Forgetting is Easier than You Think

Online Curvature-Aware Replay: Leveraging $\mathbf{2^{nd}}$ Order Information for Online Continual Learning

Lightweight-Mark: Rethinking Deep Learning-Based Watermarking

On Zero-Initialized Attention: Optimal Prompt and Gating Factor Estimation

Incorporating Arbitrary Matrix Group Equivariance into KANs

RollingQ: Reviving the Cooperation Dynamics in Multimodal Transformer

Position: The Categorization of Race in ML is a Flawed Premise

Position: Certified Robustness Does Not (Yet) Imply Model Security

Position: AI Safety Must Embrace an Antifragile Perspective

Position: Constants are Critical in Regret Bounds for Reinforcement Learning

Position: A Theory of Deep Learning Must Include Compositional Sparsity

Position: Lifetime tuning is incompatible with continual reinforcement learning

Position: Spectral GNNs Rely Less on Graph Fourier Basis than Conceived

Position: Language model developers should report train-test overlap

Position: Not All Explanations for Deep Learning Phenomena Are Equally Valuable

Position: Probabilistic Modelling is Sufficient for Causal Inference

Position: AI Scaling: From Up to Down and Out

Position: Future Research and Challenges Remain Towards AI for Software Engineering

Position: Scaling LLM Agents Requires Asymptotic Analysis with LLM Primitives

One Image is Worth a Thousand Words: A Usability Preservable Text-Image Collaborative Erasing Framework

ExpProof : Operationalizing Explanations for Confidential Models with ZKPs

Test-Time Multimodal Backdoor Detection by Contrastive Prompting

A Checks-and-Balances Framework for Context-Aware Ethical AI Alignment

Not All Wrong is Bad: Using Adversarial Examples for Unlearning

Compositional Condition Question Answering in Tabular Understanding

ShieldAgent: Shielding Agents via Verifiable Safety Policy Reasoning

Leveraging Randomness in Model and Data Partitioning for Privacy Amplification

Distributed Differentially Private Data Analytics via Secure Sketching

Identifying and Understanding Cross-Class Features in Adversarial Training

MTL-UE: Learning to Learn Nothing for Multi-Task Learning

The Jailbreak Tax: How Useful are Your Jailbreak Outputs?

The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret

STD-FD: Spatio-Temporal Distribution Fitting Deviation for AIGC Forgery Identification

It's My Data Too: Private ML for Datasets with Multi-User Training Examples

Differentially Private Federated $k$-Means Clustering with Server-Side Data

PCEvolve: Private Contrastive Evolution for Synthetic Dataset Generation via Few-Shot Private Data and Generative APIs

An Efficient Private GPT Never Autoregressively Decodes

Synthetic Face Datasets Generation via Latent Space Exploration from Brownian Identity Diffusion

(ends 1:30 PM)

Mentorship:

Science Communication 101: How to write an elevator pitch for your research

(ends 12:00 PM)

1 p.m.

Break:

Hosted Grab and Go Lunch

(ends 2:00 PM)

2 p.m.

Invited Talk:

Generative AI's Collision with Copyright Law

Pamela Samuelson

(ends 3:00 PM)

3:30 p.m.

Oral 2A Diffusion Models [3:30-4:30]

Orals 3:30-4:30

[3:30] DeFoG: Discrete Flow Matching for Graph Generation

[3:45] MGD$^3$ : Mode-Guided Dataset Distillation using Diffusion Models

[4:00] Inductive Moment Matching

[4:15] Train for the Worst, Plan for the Best: Understanding Token Ordering in Masked Diffusions

(ends 4:30 PM)

Oral 2B Positions: AI Regulation and Safety [3:30-4:30]

Orals 3:30-4:30

[3:30] Position: Generative AI Regulation Can Learn from Social Media Regulation

[3:45] Position: Current Model Licensing Practices are Dragging Us into a Quagmire of Legal Noncompliance

[4:00] Position: AI Agents Need Authenticated Delegation

[4:15] Position: AI Safety should prioritize the Future of Work

(ends 4:30 PM)

Oral 2C Reinforcement Learning [3:30-4:30]

Orals 3:30-4:30

[3:30] Controlling Underestimation Bias in Constrained Reinforcement Learning for Safe Exploration

[3:45] Temporal Difference Flows

[4:00] Network Sparsity Unlocks the Scaling Potential of Deep Reinforcement Learning

[4:15] Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination

(ends 4:30 PM)

Oral 2D Efficient ML [3:30-4:30]

Orals 3:30-4:30

[3:30] AdaSplash: Adaptive Sparse Flash Attention

[3:45] Accelerating LLM Inference with Lossless Speculative Decoding Algorithms for Heterogeneous Vocabularies

[4:00] ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features

[4:15] Mixture of Lookup Experts

(ends 4:30 PM)

Oral 2E Optimal Transport [3:30-4:30]

Orals 3:30-4:30

[3:30] Hierarchical Refinement: Optimal Transport to Infinity and Beyond

[3:45] Fully Dynamic Euclidean Bi-Chromatic Matching in Sublinear Update Time

[4:00] Flowing Datasets with Wasserstein over Wasserstein Gradient Flows

[4:15] Addressing Misspecification in Simulation-based Inference through Data-driven Calibration

(ends 4:30 PM)

4:30 p.m.

Break:

Coffee Break

(ends 5:30 PM)

Poster Session 2 West [4:30-7:00]

Posters 4:30-7:00

EGPlace: An Efficient Macro Placement Method via Evolutionary Search with Greedy Repositioning Guided Mutation

Depth Degeneracy in Neural Networks: Vanishing Angles in Fully Connected ReLU Networks on Initialization

Exact Recovery of Sparse Binary Vectors from Generalized Linear Measurements

How Distributed Collaboration Influences the Diffusion Model Training? A Theoretical Perspective

The Price of Linear Time: Error Analysis of Structured Kernel Interpolation

The Underlying Universal Statistical Structure of Natural Datasets

A Reductions Approach to Risk-Sensitive Reinforcement Learning with Optimized Certainty Equivalents

Towards Theoretical Understanding of Sequential Decision Making with Preference Feedback

Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment

Cradle: Empowering Foundation Agents towards General Computer Control

The Number of Trials Matters in Infinite-Horizon General-Utility Markov Decision Processes

Near-Optimal Sample Complexity for MDPs via Anchoring

Low-Rank Thinning

PAC-Bayes Analysis for Recalibration in Classification

When to Forget? Complexity Trade-offs in Machine Unlearning

Optimal Decision Tree Pruning Revisited: Algorithms and Complexity

Learning Minimum-Size BDDs: Towards Efficient Exact Algorithms

Principal-Agent Bandit Games with Self-Interested and Exploratory Learning Agents

Leveraging Offline Data in Linear Latent Contextual Bandits

Contextual Bandits for Unbounded Context Distributions

Exploiting Curvature in Online Convex Optimization with Delayed Feedback

Agent-as-a-Judge: Evaluate Agents with Agents

StealthInk: A Multi-bit and Stealthy Watermark for Large Language Models

What can large language models do for sustainable food?

MultiPDENet: PDE-embedded Learning with Multi-time-stepping for Accelerated Flow Simulation

Perceptually Constrained Precipitation Nowcasting Model

DiffMS: Diffusion Generation of Molecules Conditioned on Mass Spectra

Physics-Informed Generative Modeling of Wireless Channels

Diagonal Symmetrization of Neural Network Solvers for the Many-Electron Schrödinger Equation

Of Mice and Machines: A Comparison of Learning Between Real World Mice and RL Agents

Context-Informed Neural ODEs Unexpectedly Identify Broken Symmetries: Insights from the Poincaré–Hopf Theorem

KinDEL: DNA-Encoded Library Dataset for Kinase Inhibitors

Wyckoff Transformer: Generation of Symmetric Crystals

Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training

Elucidating the Design Space of Multimodal Protein Language Models

LDMol: A Text-to-Molecule Diffusion Model with Structurally Informative Latent Space Surpasses AR Models

RISE: Radius of Influence based Subgraph Extraction for 3D Molecular Graph Explanation

Neural Graph Matching Improves Retrieval Augmented Generation in Molecular Machine Learning

Steering Protein Language Models

On the Query Complexity of Verifier-Assisted Language Generation

Pretraining Generative Flow Networks with Inexpensive Rewards for Molecular Graph Generation

Exploring Vision Semantic Prompt for Efficient Point Cloud Understanding

SITCOM: Step-wise Triple-Consistent Diffusion Sampling For Inverse Problems

Control and Realism: Best of Both Worlds in Layout-to-Image without Training

Learning Adaptive Lighting via Channel-Aware Guidance

Fine-Grained Captioning of Long Videos through Scene Graph Consolidation

ReFocus: Visual Editing as a Chain of Thought for Structured Image Understanding

TUMTraf VideoQA: Dataset and Benchmark for Unified Spatio-Temporal Video Understanding in Traffic Scenes

Ca2-VDM: Efficient Autoregressive Video Diffusion Model with Causal Generation and Cache Sharing

BAME: Block-Aware Mask Evolution for Efficient N:M Sparse Training

DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization

Generalizable Multi-Camera 3D Object Detection from a Single Source via Fourier Cross-View Learning

Balancing Preservation and Modification: A Region and Semantic Aware Metric for Instruction-Based Image Editing

HarmoniCa: Harmonizing Training and Inference for Better Feature Caching in Diffusion Transformer Acceleration

Exploring Invariance in Images through One-way Wave Equations

LRA-QViT: Integrating Low-Rank Approximation and Quantization for Robust and Efficient Vision Transformers

Smoothed Preference Optimization via ReNoise Inversion for Aligning Diffusion Models with Varied Human Preferences

Flopping for FLOPs: Leveraging Equivariance for Computational Efficiency

More Than Meets the Eye: Enhancing Multi-Object Tracking Even with Prolonged Occlusions

GenZSL: Generative Zero-Shot Learning Via Inductive Variational Autoencoder

Segment Anyword: Mask Prompt Inversion for Open-Set Grounded Segmentation

Balanced Learning for Domain Adaptive Semantic Segmentation

CSV-Occ: Fusing Multi-frame Alignment for Occupancy Prediction with Temporal Cross State Space Model and Central Voting Mechanism

Understanding Complexity in VideoQA via Visual Program Generation

DVI:A Derivative-based Vision Network for INR

RealRAG: Retrieval-augmented Realistic Image Generation via Self-reflective Contrastive Learning

PiD: Generalized AI-Generated Images Detection with Pixelwise Decomposition Residuals

Perceptual-GS: Scene-adaptive Perceptual Densification for Gaussian Splatting

WorldSimBench: Towards Video Generation Models as World Simulators

FeatSharp: Your Vision Model Features, Sharper

GeoPixel: Pixel Grounding Large Multimodal Model in Remote Sensing

EFDTR: Learnable Elliptical Fourier Descriptor Transformer for Instance Segmentation

TimeBase: The Power of Minimalism in Efficient Long-term Time Series Forecasting

Separating Knowledge and Perception with Procedural Data

Semantics-aware Test-time Adaptation for 3D Human Pose Estimation

Are High-Quality AI-Generated Images More Difficult for Models to Detect?

Position: Democratic AI is Possible. The Democracy Levels Framework Shows How It Might Work.

Few-Shot Learner Generalizes Across AI-Generated Image Detection

CUPS: Improving Human Pose-Shape Estimators with Conformalized Deep Uncertainty

Contrastive Visual Data Augmentation

Think Twice, Act Once: A Co-Evolution Framework of LLM and RL for Large-Scale Decision Making

Causal Invariance-aware Augmentation for Brain Graph Contrastive Learning

PyTDC: A multimodal machine learning training, evaluation, and inference platform for biomedical foundation models

EmoGrowth: Incremental Multi-label Emotion Decoding with Augmented Emotional Relation Graph

ADIOS: Antibody Development via Opponent Shaping

Modalities Contribute Unequally: Enhancing Medical Multi-modal Learning through Adaptive Modality Token Re-balancing

Maximum Entropy Reinforcement Learning with Diffusion Policy

Breaking Silos: Adaptive Model Fusion Unlocks Better Time Series Forecasting

Reinforcement Learning Control of a Physical Robot Device for Assisted Human Walking without a Simulator

Empowering World Models with Reflection for Embodied Video Prediction

Hierarchical Equivariant Policy via Frame Transfer

VIP: Vision Instructed Pre-training for Robotic Manipulation

Test-Time Adaptation for Online Vision-Language Navigation with Feedback-based Reinforcement Learning

TS-SNN: Temporal Shift Module for Spiking Neural Networks

Discovering Symbolic Cognitive Models from Human and Animal Behavior

Hybrid Spiking Vision Transformer for Object Detection with Event Cameras

Neural Representational Consistency Emerges from Probabilistic Neural-Behavioral Representation Alignment

Accurate Identification of Communication Between Multiple Interacting Neural Populations

Efficient ANN-SNN Conversion with Error Compensation Learning

MindAligner: Explicit Brain Functional Alignment for Cross-Subject Visual Decoding from Limited fMRI Data

Agent Reviewers: Domain-specific Multimodal Agents with Shared Memory for Paper Review

ArrayDPS: Unsupervised Blind Speech Separation with a Diffusion Prior

PertEval-scFM: Benchmarking Single-Cell Foundation Models for Perturbation Effect Prediction

SPACE: Your Genomic Profile Predictor is a Powerful DNA Foundation Model

Protein Structure Tokenization: Benchmarking and New Recipe

BoxLM: Unifying Structures and Semantics of Medical Concepts for Diagnosis Prediction in Healthcare

Drug-TTA: Test-Time Adaptation for Drug Virtual Screening via Multi-task Meta-Auxiliary Learning

CLIMB: Data Foundations for Large Scale Multimodal Clinical Foundation Models

Optimal Information Retention for Time-Series Explanations

Supervised Contrastive Learning from Weakly-Labeled Audio Segments for Musical Version Matching

FLAM: Frame-Wise Language-Audio Modeling

The Case for Learned Provenance-based System Behavior Baseline

Enhancing Graph Contrastive Learning for Protein Graphs from Perspective of Invariance

Fast and Low-Cost Genomic Foundation Models via Outlier Removal

Analytical Lyapunov Function Discovery: An RL-based Generative Approach

OptMATH: A Scalable Bidirectional Data Synthesis Framework for Optimization Modeling

Stochastic Deep Restoration Priors for Imaging Inverse Problems

ROS: A GNN-based Relax-Optimize-and-Sample Framework for Max-$k$-Cut Problems

An in depth look at the Procrustes-Wasserstein distance: properties and barycenters

MVA: Linear Attention with High-order Query-Keys Integration and Multi-level Vocabulary Decomposition

Safe-EF: Error Feedback for Non-smooth Constrained Optimization

Causality Inspired Federated Learning for OOD Generalization

Beyond Communication Overhead: A Multilevel Monte Carlo Approach for Mitigating Compression Bias in Distributed Learning

Efficient Optimization with Orthogonality Constraint: a Randomized Riemannian Submanifold Method

Attention-Only Transformers via Unrolled Subspace Denoising

Recommendations with Sparse Comparison Data: Provably Fast Convergence for Nonconvex Matrix Factorization

Clipping Improves Adam-Norm and AdaGrad-Norm when the Noise Is Heavy-Tailed

Clipped SGD Algorithms for Performative Prediction: Tight Bounds for Stochastic Bias and Remedies

Hyperspherical Normalization for Scalable Deep Reinforcement Learning

Enhancing Rating-Based Reinforcement Learning to Effectively Leverage Feedback from Large Vision-Language Models

Gradual Transition from Bellman Optimality Operator to Bellman Operator in Online Reinforcement Learning

Constrained Exploitability Descent: An Offline Reinforcement Learning Method for Finding Mixed-Strategy Nash Equilibrium

Q-Supervised Contrastive Representation: A State Decoupling Framework for Safe Offline Reinforcement Learning

CERTAIN: Context Uncertainty-aware One-Shot Adaptation for Context-based Offline Meta Reinforcement Learning

Habitizing Diffusion Planning for Efficient and Effective Decision Making

Off-Policy Evaluation under Nonignorable Missing Data

Strategic Planning: A Top-Down Approach to Option Generation

Action-Dependent Optimality-Preserving Reward Shaping

Task-Agnostic Pre-training and Task-Guided Fine-tuning for Versatile Diffusion Planner

Wasserstein Policy Optimization

Controlling Underestimation Bias in Constrained Reinforcement Learning for Safe Exploration

Temporal Difference Flows

Learning Fused State Representations for Control from Multi-View Observations

Learning Policy Committees for Effective Personalization in MDPs with Diverse Tasks

Model-Based Exploration in Monitored Markov Decision Processes

ADDQ: Adaptive distributional double Q-learning

Zero-Shot Offline Imitation Learning via Optimal Transport

Gap-Dependent Bounds for Federated $Q$-Learning

Neural Event-Triggered Control with Optimal Scheduling

Understanding High-Dimensional Bayesian Optimization

Network Sparsity Unlocks the Scaling Potential of Deep Reinforcement Learning

Policy-labeled Preference Learning: Is Preference Enough for RLHF?

Reinforcement Learning with Adaptive Reward Modeling for Expensive-to-Evaluate Systems

Sliding Puzzles Gym: A Scalable Benchmark for State Representation in Visual Reinforcement Learning

IL-SOAR : Imitation Learning with Soft Optimistic Actor cRitic

Provably Efficient Exploration in Inverse Constrained Reinforcement Learning

Wolfpack Adversarial Attack for Robust Multi-Agent Reinforcement Learning

Learning Progress Driven Multi-Agent Curriculum

Distributionally Robust Multi-Agent Reinforcement Learning for Dynamic Chute Mapping

Graph Diffusion for Robust Multi-Agent Coordination

Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration

HYGMA: Hypergraph Coordination Networks with Dynamic Grouping for Multi-Agent Reinforcement Learning

Reidentify: Context-Aware Identity Generation for Contextual Multi-Agent Reinforcement Learning

Cross-environment Cooperation Enables Zero-shot Multi-agent Coordination

Improving Transformer World Models for Data-Efficient RL

Sample Complexity of Distributionally Robust Off-Dynamics Reinforcement Learning with Online Interaction

Navigating the Social Welfare Frontier: Portfolios for Multi-objective Reinforcement Learning

Flow-based Domain Randomization for Learning and Sequencing Robotic Skills

Monte-Carlo Tree Search with Uncertainty Propagation via Optimal Transport

Subgoal-Guided Policy Heuristic Search with Learned Subgoals

Online Robust Reinforcement Learning Through Monte-Carlo Planning

DiLQR: Differentiable Iterative Linear Quadratic Regulator via Implicit Differentiation

Convex Markov Games: A New Frontier for Multi-Agent Reinforcement Learning

Beyond Self-Interest: How Group Strategies Reshape Content Creation in Recommendation Platforms?

Reducing Variance of Stochastic Optimization for Approximating Nash Equilibria in Normal-Form Games

Function Encoders: A Principled Approach to Transfer Learning in Hilbert Spaces

Permutation Equivariant Neural Networks for Symmetric Tensors

Benefits of Early Stopping in Gradient Descent for Overparameterized Logistic Regression

A Rescaling-Invariant Lipschitz Bound Based on Path-Metrics for Modern ReLU Network Parameterizations

Deep Ridgelet Transform and Unified Universality Theorem for Deep and Shallow Joint-Group-Equivariant Machines

Eigen Analysis of Conjugate Kernel and Neural Tangent Kernel

Towards characterizing the value of edge embeddings in Graph Neural Networks

Optimal Task Order for Continual Learning of Multiple Tasks

Feature learning from non-Gaussian inputs: the case of Independent Component Analysis in high dimensions

Algorithms with Calibrated Machine Learning Predictions

On the Similarities of Embeddings in Contrastive Learning

Measuring Diversity: Axioms and Challenges

World Model Implanting for Test-time Adaptation of Embodied Agents

Goal-Oriented Skill Abstraction for Offline Multi-Task Reinforcement Learning

Benchmarking Quantum Reinforcement Learning

Preference Controllable Reinforcement Learning with Advanced Multi-Objective Optimization

An Analysis of Quantile Temporal-Difference Learning

The Empirical Mean is Minimax Optimal for Local Glivenko-Cantelli

Design Considerations in Offline Preference-based RL

Adversarial Robust Generalization of Graph Neural Networks

Ehrenfeucht-Haussler Rank and Chain of Thought

Transfer Q-Learning with Composite MDP Structures

LoRA-One: One-Step Full Gradient Could Suffice for Fine-Tuning Large Language Models, Provably and Efficiently

Weak-to-Strong Generalization Even in Random Feature Networks, Provably

Generalization and Robustness of the Tilted Empirical Risk

Stability and Generalization Analysis of Decentralized SGD: Sharper Bounds Beyond Lipschitzness and Smoothness

Understanding Generalization in Quantum Machine Learning with Margins

Generalization Analysis for Supervised Contrastive Representation Learning under Non-IID Settings

A Unified Framework for Generalization Error Analysis of Learning with Arbitrary Discrete Weak Features

On the Convergence of Continuous Single-timescale Actor-critic

Discrepancies are Virtue: Weak-to-Strong Generalization through Lens of Intrinsic Dimension

Retraining with Predicted Hard Labels Provably Increases Model Accuracy

Revisiting Differentially Private Algorithms for Decentralized Online Learning

Near-Optimal Consistency-Robustness Trade-Offs for Learning-Augmented Online Knapsack Problems

Improved Online Confidence Bounds for Multinomial Logistic Bandits

A Parametric Contextual Online Learning Theory of Brokerage

Feasible Action Search for Bandit Linear Programs via Thompson Sampling

Tokenized Bandit for LLM Decoding and Alignment

Online Linear Classification with Massart Noise

(ends 7:00 PM)

Poster Session 2 East [4:30-7:00]

Posters 4:30-7:00

Uncertainty-Based Extensible Codebook for Discrete Federated Learning in Heterogeneous Data Silos

DRAG: Data Reconstruction Attack using Guided Diffusion

Approximate Differential Privacy of the $\ell_2$ Mechanism

Contrastive Private Data Synthesis via Weighted Multi-PLM Fusion

Empirical Privacy Variance

Textual Unlearning Gives a False Sense of Unlearning

Private Model Personalization Revisited

Equivariant Polynomial Functional Networks

Theoretically Unmasking Inference Attacks Against LDP-Protected Clients in Federated Vision Models

Scalable Private Partition Selection via Adaptive Weighting

Subgroups Matter for Robust Bias Mitigation

Multiaccuracy and Multicalibration via Proxy Groups

Dissecting Submission Limit in Desk-Rejections: A Mathematical Analysis of Fairness in AI Conference Policies

SPRI: Aligning Large Language Models with Context-Situated Principles

Massive Values in Self-Attention Modules are the Key to Contextual Knowledge Understanding

SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders

Constrain Alignment with Sparse Autoencoders

Representation Shattering in Transformers: A Synthetic Study with Knowledge Editing

Invariance Makes LLM Unlearning Resilient Even to Unanticipated Downstream Fine-Tuning

Inducing, Detecting and Characterising Neural Modules: A Pipeline for Functional Interpretability in Reinforcement Learning

A Unified Comparative Study with Generalized Conformity Scores for Multi-Output Conformal Regression

SEMU: Singular Value Decomposition for Efficient Machine Unlearning

From Individual Experience to Collective Evidence: A Reporting-Based Framework for Identifying Systemic Harms

One Wave To Explain Them All: A Unifying Perspective On Feature Attribution

Explaining, Fast and Slow: Abstraction and Refinement of Provable Explanations

Algorithmic Recourse for Long-Term Improvement

Textural or Textual: How Vision-Language Models Read Text in Images

A Non-Asymptotic Convergent Analysis for Scored-Based Graph Generative Model via a System of Stochastic Differential Equations

Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment

GEFA: A General Feature Attribution Framework Using Proxy Gradient Estimation

Flowing Datasets with Wasserstein over Wasserstein Gradient Flows

Adaptive Learn-then-Test: Statistically Valid and Efficient Hyperparameter Selection

e-GAI: e-value-based Generalized $\alpha$-Investing for Online False Discovery Rate Control

Flexible Tails for Normalizing Flows

Multilayer Matrix Factorization via Dimension-Reducing Diffusion Variational Inference

Synonymous Variational Inference for Perceptual Image Compression

An Efficient Search-and-Score Algorithm for Ancestral Graphs using Multivariate Information Scores for Complex Non-linear and Categorical Data

The Polynomial Stein Discrepancy for Assessing Moment Convergence

Importance Corrected Neural JKO Sampling

A Generic Family of Graphical Models: Diversity, Efficiency, and Heterogeneity

G-Sim: Generative Simulations with Large Language Models and Gradient-Free Calibration

Improving the Statistical Efficiency of Cross-Conformal Prediction

Gridded Transformer Neural Processes for Spatio-Temporal Data

Rethinking Aleatoric and Epistemic Uncertainty

Addressing Misspecification in Simulation-based Inference through Data-driven Calibration

BARNN: A Bayesian Autoregressive and Recurrent Neural Network

A Bayesian Model Selection Criterion for Selecting Pretraining Checkpoints

Optimization Proxies using Limited Labeled Data and Training Time -- A Semi-Supervised Bayesian Neural Network Approach

Step-DAD: Semi-Amortized Policy-Based Bayesian Experimental Design

Finite-Time Analysis of Discrete-Time Stochastic Interpolants

FedSMU: Communication-Efficient and Generalization-Enhanced Federated Learning through Symbolic Model Updates

Flexible, Efficient, and Stable Adversarial Attacks on Machine Unlearning

Fully Dynamic Euclidean Bi-Chromatic Matching in Sublinear Update Time

Towards an Explainable Comparison and Alignment of Feature Embeddings

Expressive Score-Based Priors for Distribution Matching with Geometry-Preserving Regularization

Leveraging Diffusion Model as Pseudo-Anomalous Graph Generator for Graph-Level Anomaly Detection

Neurosymbolic World Models for Sequential Decision Making

Test-time Correlation Alignment

Parameter-Efficient Fine-Tuning of State Space Models

Pareto Merging: Multi-Objective Optimization for Preference-Aware Model Merging

Balancing the Scales: A Theoretical and Algorithmic Framework for Learning from Imbalanced Data

Label Distribution Propagation-based Label Completion for Crowdsourcing

Text-to-LoRA: Instant Transformer Adaption

In-Context Learning and Occam's Razor

Improving Continual Learning Performance and Efficiency with Auxiliary Classifiers

Zero-Shot Adaptation of Parameter-Efficient Fine-Tuning in Diffusion Models

BECAME: Bayesian Continual Learning with Adaptive Model Merging

CALM: Consensus-Aware Localized Merging for Multi-Task Learning

QT-DoG: Quantization-Aware Training for Domain Generalization

Learning Time-Aware Causal Representation for Model Generalization in Evolving Domains

Stray Intrusive Outliers-Based Feature Selection on Intra-Class Asymmetric Instance Distribution or Multiple High-Density Clusters

Prediction models that learn to avoid missing values

Survival Analysis via Density Estimation

Learning Changes in Graphon Attachment Network Models

Time Series Representations with Hard-Coded Invariances

Sample Complexity of Correlation Detection in the Gaussian Wigner Model

Relational Conformal Prediction for Correlated Time Series

Complete-Tree Space Favors Data-Efficient Link Prediction

Contextures: Representations from Contexts

One Example Shown, Many Concepts Known! Counterexample-Driven Conceptual Reasoning in Mathematical LLMs

PaperBench: Evaluating AI’s Ability to Replicate AI Research

MathConstruct: Challenging LLM Reasoning with Constructive Proofs

RBench: Graduate-level Multi-disciplinary Benchmarks for LLM & MLLM Complex Reasoning Evaluation

HiRemate: Hierarchical Approach for Efficient Re-materialization of Neural Networks

BaWA: Automatic Optimizing Pruning Metric for Large Language Models with Balanced Weight and Activation

Fusing Reward and Dueling Feedback in Stochastic Bandits

Stochastic Online Conformal Prediction with Semi-Bandit Feedback

MOGIC: Metadata-infused Oracle Guidance for Improved Extreme Classification

Rethinking Point Cloud Data Augmentation: Topologically Consistent Deformation

Nonparametric Identification of Latent Concepts

Volume-Aware Distance for Robust Similarity Learning

Learning Distances from Data with Normalizing Flows and Score Matching

Model Steering: Learning with a Reference Model Improves Generalization Bounds and Scaling Laws

Scalable Attribute-Missing Graph Clustering via Neighborhood Differentiation

Automatically Identify and Rectify: Robust Deep Contrastive Multi-view Clustering in Noisy Scenarios

Adjustment for Confounding using Pre-Trained Representations

Heterogeneous Treatment Effect in Time-to-Event Outcomes: Harnessing Censored Data with Recursively Imputed Trees

AutoCATE: End-to-End, Automated Treatment Effect Estimation

Enhancing Statistical Validity and Power in Hybrid Controlled Trials: A Randomization Inference Approach with Conformal Selective Borrowing

A Meta-learner for Heterogeneous Effects in Difference-in-Differences

Hierarchical Reinforcement Learning with Targeted Causal Interventions

Linear Contextual Bandits With Interference

Extracting Rare Dependence Patterns via Adaptive Sample Reweighting

Active Treatment Effect Estimation via Limited Samples

Strong and Weak Identifiability of Optimization-based Causal Discovery in Non-linear Additive Noise Models

Continuous Bayesian Model Selection for Multivariate Causal Discovery

Fast Min-$\epsilon$ Segmented Regression using Constant-Time Segment Merging

UltraTWD: Optimizing Ultrametric Trees for Tree-Wasserstein Distance

Branches: Efficiently Seeking Optimal Sparse Decision Trees via AO*

Solving Satisfiability Modulo Counting Exactly with Probabilistic Circuits

Locality Preserving Markovian Transition for Instance Retrieval

Conformity Score Averaging for Classification

False Coverage Proportion Control for Conformal Prediction

A Variational Information Theoretic Approach to Out-of-Distribution Detection

Hierarchical Refinement: Optimal Transport to Infinity and Beyond

Deep Sturm–Liouville: From Sample-Based to 1D Regularization with Learnable Orthogonal Basis Functions

A Certified Unlearning Approach without Access to Source Data

One Arrow, Two Hawks: Sharpness-aware Minimization for Federated Learning via Global Model Trajectory

Does One-shot Give the Best Shot? Mitigating Model Inconsistency in One-shot Federated Learning

Prior Knowledge Guided Neural Architecture Generation

Stochastic Encodings for Active Feature Acquisition

FastCAV: Efficient Computation of Concept Activation Vectors for Explaining Deep Neural Networks

Curvature Enhanced Data Augmentation for Regression

On Temperature Scaling and Conformal Prediction of Deep Classifiers

The Hidden Joules: Evaluating the Energy Consumption of Vision Backbones for Progress Towards More Efficient Model Inference

The Butterfly Effect: Neural Network Training Trajectories Are Highly Sensitive to Initial Conditions

Floating-Point Neural Networks Can Represent Almost All Floating-Point Functions

Adaptive kernel predictors from feature-learning infinite limits of neural networks

On the Local Complexity of Linear Regions in Deep ReLU Networks

Fishers for Free? Approximating the Fisher Information Matrix by Recycling the Squared Gradient Accumulator

Optimization for Neural Operators can Benefit from Width

IMTS is Worth Time $\times$ Channel Patches: Visual Masked Autoencoders for Irregular Multivariate Time Series Prediction

Probably Approximately Global Robustness Certification

Adapting to Evolving Adversaries with Regularized Continual Robust Training

SMART-PC: Skeletal Model Adaptation for Robust Test-Time Training in Point Clouds

CTBench: A Library and Benchmark for Certified Training

Targeted Unlearning with Single Layer Unlearning Gradient

An Augmentation-Aware Theory for Self-Supervised Contrastive Learning

Patch-wise Structural Loss for Time Series Forecasting

AEQA-NAT : Adaptive End-to-end Quantization Alignment Training Framework for Non-autoregressive Machine Translation

Discovering Physics Laws of Dynamical Systems via Invariant Function Learning

Bayesian Basis Function Approximation for Scalable Gaussian Process Priors in Deep Generative Models

Non-stationary Diffusion For Probabilistic Time Series Forecasting

When Will It Fail?: Anomaly to Prompt for Forecasting Future Anomalies in Time Series

Towards a General Time Series Forecasting Model with Unified Representation and Adaptive Transfer

Unveiling AI's Blind Spots: An Oracle for In-Domain, Out-of-Domain, and Adversarial Errors

The Devil Is in the Details: Tackling Unimodal Spurious Correlations for Generalizable Multimodal Reward Models

AdvAgent: Controllable Blackbox Red-teaming on Web Agents

Ranked Entropy Minimization for Continual Test-Time Adaptation

Mahalanobis++: Improving OOD Detection via Feature Normalization

Efficient Robust Conformal Prediction via Lipschitz-Bounded Networks

Self-supervised Adversarial Purification for Graph Neural Networks

Beyond One-Hot Labels: Semantic Mixing for Model Calibration

$\texttt{I$^2$MoE}$: Interpretable Multimodal Interaction-aware Mixture-of-Experts

Multi-band Frequency Reconstruction for Neural Psychoacoustic Coding

Canonical Rank Adaptation: An Efficient Fine-Tuning Strategy for Vision Transformers

CPCF: A Cross-Prompt Contrastive Framework for Referring Multimodal Large Language Models

NoLiMa: Long-Context Evaluation Beyond Literal Matching

Defending LVLMs Against Vision Attacks Through Partial-Perception Supervision

Teaching Physical Awareness to LLMs through Sounds

IBCircuit: Towards Holistic Circuit Discovery with Information Bottleneck

Accelerating Unbiased LLM Evaluation via Synthetic Feedback

WikiBigEdit: Understanding the Limits of Lifelong Knowledge Editing in LLMs

(How) Do Language Models Track State?

PDE-Controller: LLMs for Autoformalization and Reasoning of PDEs

Reinforce LLM Reasoning through Multi-Agent Reflection

Premise-Augmented Reasoning Chains Improve Error Identification in Math reasoning with LLMs

PoisonBench: Assessing Language Model Vulnerability to Poisoned Preference Data

Attention Mechanisms Perspective: Exploring LLM Processing of Graph-Structured Data

Vulnerability-Aware Alignment: Mitigating Uneven Forgetting in Harmful Fine-Tuning

Idiosyncrasies in Large Language Models

Active Reward Modeling: Adaptive Preference Labeling for Large Language Model Alignment

Understanding Chain-of-Thought in LLMs through Information Theory

Improving Model Alignment Through Collective Intelligence of Open-Source Models

KV Shifting Attention Enhances Language Modeling

DA-KD: Difficulty-Aware Knowledge Distillation for Efficient Large Language Models

Puzzle: Distillation-Based NAS for Inference-Optimized LLMs

Analyze Feature Flow to Enhance Interpretation and Steering in Language Models

Steer LLM Latents for Hallucination Detection

Token Signature: Predicting Chain-of-Thought Gains with Token Decoding Feature in Large Language Models

Maximizing Intermediate Checkpoint Value in LLM Pretraining with Bayesian Optimization

Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search

Preference Learning for AI Alignment: a Causal Perspective

FloE: On-the-Fly MoE Inference on Memory-constrained GPU

Olica: Efficient Structured Pruning of Large Language Models without Retraining

Token Assorted: Mixing Latent and Text Tokens for Improved Language Model Reasoning

To Steer or Not to Steer? Mechanistic Error Reduction with Abstention for Language Models

Revealing Weaknesses in Text Watermarking Through Self-Information Rewrite Attacks

Unlocking Post-hoc Dataset Inference with Synthetic Data

ProSec: Fortifying Code LLMs with Proactive Security Alignment

Lexico: Extreme KV Cache Compression via Sparse Coding over Universal Dictionaries

PIPA: Preference Alignment as Prior-Informed Statistical Estimation

Regress, Don't Guess: A Regression-like Loss on Number Tokens for Language Models

Evolving Prompts In-Context: An Open-ended, Self-replicating Perspective

Unnatural Languages Are Not Bugs but Features for LLMs

LongRoPE2: Near-Lossless LLM Context Window Scaling

Potemkin Understanding in Large Language Models

A Tale of Two Structures: Do LLMs Capture the Fractal Complexity of Language?

Reflection-Window Decoding: Text Generation with Selective Refinement

MxMoE: Mixed-precision Quantization for MoE with Accuracy and Performance Co-Design

Generalizing from SIMPLE to HARD Visual Reasoning: Can We Mitigate Modality Imbalance in VLMs?

PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling

SNS-Bench: Defining, Building, and Assessing Capabilities of Large Language Models in Social Networking Services

CommVQ: Commutative Vector Quantization for KV Cache Compression

Evaluating LLMs Across Multi-Cognitive Levels: From Medical Knowledge Mastery to Scenario-Based Problem Solving

Efficient Graph Continual Learning via Lightweight Graph Neural Tangent Kernels-based Dataset Distillation

The Price of Freedom: Exploring Expressivity and Runtime Tradeoffs in Equivariant Tensor Products

Test-Time Graph Neural Dataset Search With Generative Projection

CSG-ODE: ControlSynth Graph ODE For Modeling Complex Evolution of Dynamic Graphs

A Manifold Perspective on the Statistical Generalization of Graph Neural Networks

Structure Is All You Need: Structural Representation Learning on Hyper-Relational Knowledge Graphs

Self-Improving Language Models for Evolutionary Program Synthesis: A Case Study on ARC-AGI

Mixture of Lookup Experts

CRANE: Reasoning with constrained LLM generation

Accelerating LLM Inference with Lossless Speculative Decoding Algorithms for Heterogeneous Vocabularies

What Makes In-context Learning Effective for Mathematical Reasoning

any4: Learned 4-bit Numeric Representation for LLMs

CurvGAD: Leveraging Curvature for Enhanced Graph Anomaly Detection

Implicit Subgraph Neural Network

Enhancing the Influence of Labels on Unlabeled Nodes in Graph Convolutional Networks

How Much Can Transfer? BRIDGE: Bounded Multi-Domain Graph Foundation Model with Generalization Guarantees

Implicit degree bias in the link prediction task

Redundancy Undermines the Trustworthiness of Self-Interpretable GNNs

Learn Beneficial Noise as Graph Augmentation

Toward Data-centric Directed Graph Learning: An Entropy-driven Approach

SCENIR: Visual Semantic Clarity through Unsupervised Scene Graph Retrieval

Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan

Ultra-Resolution Adaptation with Ease

Geometry Informed Tokenization of Molecules for Language Model Generation

Kinetic Langevin Diffusion for Crystalline Materials Generation

SADA: Stability-guided Adaptive Diffusion Acceleration

ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features

MGD$^3$ : Mode-Guided Dataset Distillation using Diffusion Models

Learnings from Scaling Visual Tokenizers for Reconstruction and Generation

DeFoG: Discrete Flow Matching for Graph Generation

Understanding and Mitigating Memorization in Diffusion Models for Tabular Data

Boost-and-Skip: A Simple Guidance-Free Diffusion for Minority Generation

Scaling Laws for Pre-training Agents and World Models

Understanding and Mitigating Memorization in Generative Models via Sharpness of Probability Landscapes

Train for the Worst, Plan for the Best: Understanding Token Ordering in Masked Diffusions

FlexTok: Resampling Images into 1D Token Sequences of Flexible Length

Devil is in the Details: Density Guidance for Detail-Aware Generation with Flow Models

Accelerated Diffusion Models via Speculative Sampling

Inverse Bridge Matching Distillation

Noise Conditional Variational Score Distillation

D-Fusion: Direct Preference Optimization for Aligning Diffusion Models with Visually Consistent Samples

Shielded Diffusion: Generating Novel and Diverse Images using Sparse Repellency

Efficiently Access Diffusion Fisher: Within the Outer Product Span Space

Compressed Image Generation with Denoising Diffusion Codebook Models

Identifiable Object Representations under Spatial Ambiguities

Beyond Log-Concavity and Score Regularity: Improved Convergence Bounds for Score-Based Generative Models in W2-distance

LGDM: Latent Guidance in Diffusion Models for Perceptual Evaluations

Concept Reachability in Diffusion Models: Beyond Dataset Constraints

Inductive Moment Matching

Beyond Task-Specific Reasoning: A Unified Conditional Generative Framework for Abstract Visual Reasoning

Skrr: Skip and Re-use Text Encoder Layers for Memory Efficient Text-to-Image Generation

Orthus: Autoregressive Interleaved Image-Text Generation with Modality-Specific Heads

Portable Reward Tuning: Towards Reusable Fine-Tuning across Different Pretrained Models

In-Context Deep Learning via Transformer Models

The Double-Ellipsoid Geometry of CLIP

Sundial: A Family of Highly Capable Time Series Foundation Models

Are Large Brainwave Foundation Models Capable Yet ? Insights from Fine-Tuning

Whitened CLIP as a Likelihood Surrogate of Images and Captions

How Far Is Video Generation from World Model: A Physical Law Perspective

TabICL: A Tabular Foundation Model for In-Context Learning on Large Data

Score-of-Mixture Training: One-Step Generative Model Training Made Simple via Score Estimation of Mixture Distributions

Vector Grimoire: Codebook-based Shape Generation under Raster Image Supervision

Latent Thought Models with Variational Bayes Inference-Time Computation

UniDB: A Unified Diffusion Bridge Framework via Stochastic Optimal Control

LowRA: Accurate and Efficient LoRA Fine-Tuning of LLMs under 2 Bits

WildChat-50M: A Deep Dive Into the Role of Synthetic Data in Post-Training

On Understanding Attention-Based In-Context Learning for Categorical Data

Benign Overfitting in Token Selection of Attention Mechanism

AdaSplash: Adaptive Sparse Flash Attention

Star Attention: Efficient LLM Inference over Long Sequences

Sparse Video-Gen: Accelerating Video Diffusion Transformers with Spatial-Temporal Sparsity

Re-ranking Reasoning Context with Tree Search Makes Large Vision-Language Models Stronger

On-Device Collaborative Language Modeling via a Mixture of Generalists and Specialists

Beyond Low-rank Decomposition: A Shortcut Approach for Efficient On-Device Learning

SAND: One-Shot Feature Selection with Additive Noise Distortion

Unisolver: PDE-Conditional Transformers Towards Universal Neural PDE Solvers

Unlocking the Power of SAM 2 for Few-Shot Segmentation

Neural Genetic Search in Discrete Spaces

Semantic Shift Estimation via Dual-Projection and Classifier Reconstruction for Exemplar-Free Class-Incremental Learning

The Importance of Being Lazy: Scaling Limits of Continual Learning

Compression via Pre-trained Transformers: A Study on Byte-Level Multimodal Data

L3A: Label-Augmented Analytic Adaptation for Multi-Label Class Incremental Learning

Improved Algorithm for Deep Active Learning under Imbalance via Optimal Separation

Position: AI Safety should prioritize the Future of Work

Position: AI's growing due process problem

Position: Generative AI Regulation Can Learn from Social Media Regulation

Position: Current Model Licensing Practices are Dragging Us into a Quagmire of Legal Noncompliance

Position: Editing Large Language Models Poses Serious Safety Risks

Position: Machine Learning Models Have a Supply Chain Problem

Position: AI Agents Need Authenticated Delegation

Position: LLM Social Simulations Are a Promising Research Method

Position: Challenges and Future Directions of Data-Centric AI Alignment

Position: Causal Machine Learning Requires Rigorous Synthetic Experiments for Broader Adoption

Position: Formal Mathematical Reasoning—A New Frontier in AI

Position: An Empirically Grounded Identifiability Theory Will Accelerate Self Supervised Learning Research

Position: Enough of Scaling LLMs! Lets Focus on Downscaling

Position: In-House Evaluation Is Not Enough. Towards Robust Third-Party Evaluation and Flaw Disclosure for General-Purpose AI

Position: We Need An Algorithmic Understanding of Generative AI

Position: Beyond Assistance – Reimagining LLMs as Ethical and Adaptive Co-Creators in Mental Health Care

Position: Algebra Unveils Deep Learning - An Invitation to Neuroalgebraic Geometry

Position: Trustworthy AI Agents Require the Integration of Large Language Models and Formal Methods

Position: Graph Matching Systems Deserve Better Benchmarks

MixBridge: Heterogeneous Image-to-Image Backdoor Attack through Mixture of Schrödinger Bridges

Stay-Positive: A Case for Ignoring Real Image Features in Fake Image Detection

Adversaries Can Misuse Combinations of Safe Models

MemFreezing: A Novel Adversarial Attack on Temporal Graph Neural Networks under Limited Future Knowledge

GaussMarker: Robust Dual-Domain Watermark for Diffusion Models

Does Low Rank Adaptation Lead to Lower Robustness against Training-Time Attacks?

Black-Box Adversarial Attacks on LLM-Based Code Completion

Byzantine-Resilient Federated Alternating Gradient Descent and Minimization for Partly-Decoupled Low Rank Matrix Learning

Gandalf the Red: Adaptive Security for LLMs

WMarkGPT: Watermarked Image Understanding via Multimodal Large Language Models

Speak Easy: Eliciting Harmful Jailbreaks from LLMs with Simple Interactions

MP-Nav: Enhancing Data Poisoning Attacks against Multimodal Learning

Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens

Learning Classifiers That Induce Markets

Certified Unlearning for Neural Networks

Clients Collaborate: Flexible Differentially Private Federated Learning with Guaranteed Improvement of Utility-Privacy Trade-off

(ends 7:00 PM)

7 p.m.

WED 16 JUL

7:30 a.m.

Registration West

(ends 6:00 PM)

Registration East

(ends 12:00 PM)

ICML Lounge Area:

ICML Lounge Area

(ends 7:00 PM)

8:30 a.m.

Test Of Time:

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift

(ends 9:30 AM)

9:30 a.m.

Break:

Coffee Break

(ends 10:00 AM)

Exhibits

(ends 6:00 PM)

10 a.m.

Affinity Workshop:

WiML

(ends 9:00 PM)

Oral 3A Reasoning [10:00-11:00]

Orals 10:00-11:00

[10:00] Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction

[10:15] Can MLLMs Reason in Multimodality? EMMA: An Enhanced MultiModal ReAsoning Benchmark

[10:30] rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking

[10:45] VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data

(ends 11:00 AM)

Oral 3B Representations 1 [10:00-11:00]

Orals 10:00-11:00

[10:00] A Generalization Theory for Zero-Shot Prediction

[10:15] Statistical Test for Feature Selection Pipelines by Selective Inference

[10:30] Learning with Expected Signatures: Theory and Applications

[10:45] Blink of an eye: a simple theory for feature localization in generative models

(ends 11:00 AM)

Oral 3C Data-Centric ML [10:00-11:00]

Orals 10:00-11:00

[10:00] Outlier Gradient Analysis: Efficiently Identifying Detrimental Training Samples for Deep Learning Models

[10:15] Foundation Model Insights and a Multi-Model Approach for Superior Fine-Grained One-shot Subset Selection

[10:30] SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMs

[10:45] Improving the Scaling Laws of Synthetic Data with Deliberate Practice

(ends 11:00 AM)

Oral 3D Optimization [10:00-11:00]

Orals 10:00-11:00

[10:00] Nonlinearly Preconditioned Gradient Methods under Generalized Smoothness

[10:15] An Online Adaptive Sampling Algorithm for Stochastic Difference-of-convex Optimization with Time-varying Distributions

[10:30] Fundamental Bias in Inverting Random Sampling Matrices with Application to Sub-sampled Newton

[10:45] General framework for online-to-nonconvex conversion: Schedule-free SGD is also effective for nonconvex optimization

(ends 11:00 AM)

Oral 3E Causality and Domain Generalization [10:00-11:00]

Orals 10:00-11:00

[10:00] One-Step Generalization Ratio Guided Optimization for Domain Generalization

[10:15] An Improved Clique-Picking Algorithm for Counting Markov Equivalent DAGs via Super Cliques Transfer

[10:30] Polynomial-Delay MAG Listing with Novel Locally Complete Orientation Rules

[10:45] Sanity Checking Causal Representation Learning on a Simple Real-World System

(ends 11:00 AM)

11 a.m.

Poster Session 3 West [11:00-1:30]

Posters 11:00-1:30

MindLLM: A Subject-Agnostic and Versatile Model for fMRI-to-text Decoding

Tackling View-Dependent Semantics in 3D Language Gaussian Splatting

Polynomial Time Learning Augmented Algorithms for NP-hard Permutation Problems

QuEst: Enhancing Estimates of Quantile-Based Distributional Measures Using Model Predictions

Fragments to Facts: Partial-Information Fragment Inference from LLMs

Combinatorial Reinforcement Learning with Preference Feedback

Partially Observable Reinforcement Learning with Memory Traces

Multi-objective Linear Reinforcement Learning with Lexicographic Rewards

Near-optimal Regret Using Policy Optimization in Online MDPs with Aggregate Bandit Feedback

A Theoretical Justification for Asymmetric Actor-Critic Algorithms

Universal Approximation Theorem of Deep Q-Networks

AdaWorld: Learning Adaptable World Models with Latent Actions

Demystifying the Paradox of Importance Sampling with an Estimated History-Dependent Behavior Policy in Off-Policy Evaluation

Multiple-policy Evaluation via Density Estimation

Provable Zero-Shot Generalization in Offline Reinforcement Learning

Learning with Expected Signatures: Theory and Applications

Efficient Diffusion Models for Symmetric Manifolds

Diffusion Models are Secretly Exchangeable: Parallelizing DDPMs via Auto Speculation

The Limits of Tractable Marginalization

Exact Upper and Lower Bounds for the Output Distribution of Neural Networks with Random Inputs

Distributed Conformal Prediction via Message Passing

Blink of an eye: a simple theory for feature localization in generative models

Federated Oriented Learning: A Practical One-Shot Personalized Federated Learning Framework

Statistical Test for Feature Selection Pipelines by Selective Inference

Making Hard Problems Easier with Custom Data Distributions and Loss Regularization: A Case Study in Modular Arithmetic

NMA-tune: Generating Highly Designable and Dynamics Aware Protein Backbones

WGFormer: An SE(3)-Transformer Driven by Wasserstein Gradient Flows for Molecular Ground-State Conformation Prediction

MF-LAL: Drug Compound Generation Using Multi-Fidelity Latent Space Active Learning

DISCO: learning to DISCover an evolution Operator for multi-physics-agnostic prediction

PolyConf: Unlocking Polymer Conformation Generation through Hierarchical Generative Models

Empower Structure-Based Molecule Optimization with Gradient Guided Bayesian Flow Networks

DUNIA: Pixel-Sized Embeddings via Cross-Modal Alignment for Earth Observation Applications

Learning Adversarial MDPs with Stochastic Hard Constraints

Scaling Laws in Patchification: An Image Is Worth 50,176 Tokens And More

SPEX: Scaling Feature Interaction Explanations for LLMs

DS-VLM: Diffusion Supervision Vision Language Model

Multi-Modal Object Re-identification via Sparse Mixture-of-Experts

Guided Structural Inference: Leveraging Priors with Soft Gating Mechanisms

What Has a Foundation Model Found? Inductive Bias Reveals World Models

Relating Misfit to Gain in Weak-to-Strong Generalization Beyond the Squared Loss

Enhancing Spectral GNNs: From Topology and Perturbation Perspectives

Finding Wasserstein Ball Center: Efficient Algorithm and The Applications in Fairness

Neural Interpretable PDEs: Harmonizing Fourier Insights with Attention for Scalable and Interpretable Physics Discovery

QEM-Bench: Benchmarking Learning-based Quantum Error Mitigation and QEMFormer as a Multi-ranged Context Learning Baseline

Action-Minimization Meets Generative Modeling: Efficient Transition Path Sampling with the Onsager-Machlup Functional

Chaos Meets Attention: Transformers for Large-Scale Dynamical Prediction

Scalable Equilibrium Sampling with Sequential Boltzmann Generators

Symmetry-Driven Discovery of Dynamical Variables in Molecular Simulations

Learning the Electronic Hamiltonian of Large Atomic Structures

Beyond Atoms: Enhancing Molecular Pretrained Representations with 3D Space Modeling

Broadband Ground Motion Synthesis by Diffusion Model with Minimal Condition

EcoMapper: Generative Modeling for Climate-Aware Satellite Imagery

Offline Model-based Optimization for Real-World Molecular Discovery

HEAP: Hyper Extended A-PDHG Operator for Constrained High-dim PDEs

QuanONet: Quantum Neural Operator with Application to Differential Equation

Enhancing Ligand Validity and Affinity in Structure-Based Drug Design with Multi-Reward Optimization

From Thousands to Billions: 3D Visual Language Grounding via Render-Supervised Distillation from 2D VLMs

FourierMamba: Fourier Learning Integration with State Space Models for Image Deraining

FlexiReID: Adaptive Mixture of Expert for Multi-Modal Person Re-Identification

Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation

Efficient Noise Calculation in Deep Learning-based MRI Reconstructions

Probing Visual Language Priors in VLMs

SECOND: Mitigating Perceptual Hallucination in Vision-Language Models via Selective and Contrastive Decoding

Toward Robust Hyper-Detailed Image Captioning: A Multiagent Approach and Dual Evaluation Metrics for Factuality and Coverage

Beyond Entropy: Region Confidence Proxy for Wild Test-Time Adaptation

Demystifying Catastrophic Forgetting in Two-Stage Incremental Object Detector

Boosting Virtual Agent Learning and Reasoning: A Step-Wise, Multi-Dimensional, and Generalist Reward Model with Benchmark

Impossible Videos

FreeMesh: Boosting Mesh Generation with Coordinates Merging

Temporal Misalignment in ANN-SNN Conversion and its Mitigation via Probabilistic Spiking Neurons

Adaptive Multi-prompt Contrastive Network for Few-shot Out-of-distribution Detection

LOCATE 3D: Real-World Object Localization via Self-Supervised Learning in 3D

How Do Images Align and Complement LiDAR? Towards a Harmonized Multi-modal 3D Panoptic Segmentation

Geometric Feature Embedding for Effective 3D Few-Shot Class Incremental Learning

3D-LMVIC: Learning-based Multi-View Image Compression with 3D Gaussian Geometric Priors

Enhancing Foundation Models with Federated Domain Knowledge Infusion

Capturing Temporal Dynamics in Large-Scale Canopy Tree Height Estimation

Code-Generated Graph Representations Using Multiple LLM Agents for Material Properties Prediction

Context Matters: Query-aware Dynamic Long Sequence Modeling of Gigapixel Images

Enforcing Latent Euclidean Geometry in Single-Cell VAEs for Manifold Interpolation

Beyond Sensor Data: Foundation Models of Behavioral Data from Wearables Improve Health Predictions

Position: Supervised Classifiers Answer the Wrong Questions for OOD Detection

AutoElicit: Using Large Language Models for Expert Prior Elicitation in Predictive Modelling

MedRAX: Medical Reasoning Agent for Chest X-ray

Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching

MMedPO: Aligning Medical Vision-Language Models with Clinical-Aware Multimodal Preference Optimization

Active Learning for Efficient Discovery of Optimal Combinatorial Perturbations

A Model of Place Field Reorganization During Reward Maximization

Differential Coding for Training-Free ANN-to-SNN Conversion

Generating Hypotheses of Dynamic Causal Graphs in Neuroscience: Leveraging Generative Factor Models of Observed Time Series

Flow-field inference from neural data using deep recurrent networks

NeuroTree: Hierarchical Functional Brain Pathway Decoding for Mental Health Disorders

TTFSFormer: A TTFS-based Lossless Conversion of Spiking Transformer

Can Biologically Plausible Temporal Credit Assignment Rules Match BPTT for Neural Similarity? E-prop as an Example

SE(3)-Equivariant Diffusion Policy in Spherical Fourier Space

Three-Dimensional Trajectory Prediction with 3DMoTraj Dataset

WOMD-Reasoning: A Large-Scale Dataset for Interaction Reasoning in Driving

Hierarchical Planning for Complex Tasks with Knowledge Graph-RAG and Symbolic Verification

Latent Diffusion Planning for Imitation Learning

ABNet: Adaptive explicit-Barrier Net for Safe and Scalable Robot Learning

SeedLoRA: A Fusion Approach to Efficient LLM Fine-Tuning

Subspace Optimization for Large Language Models with Convergence Guarantees

Alberta Wells Dataset: Pinpointing Oil and Gas Wells from Satellite Imagery

HyperIV: Real-time Implied Volatility Smoothing

Dataflow-Guided Neuro-Symbolic Language Models for Type Inference

Voronoi-grid-based Pareto Front Learning and Its Application to Collaborative Federated Learning

Machines and Mathematical Mutations: Using GNNs to Characterize Quiver Mutation Classes

Physics Aware Neural Networks for Unsupervised Binding Energy Prediction

ETTA: Elucidating the Design Space of Text-to-Audio Models

Learning Initial Basis Selection for Linear Programming via Duality-Inspired Tripartite Graph Representation and Comprehensive Supervision

AUTOCIRCUIT-RL: Reinforcement Learning-Driven LLM for Automated Circuit Topology Generation

SK-VQA: Synthetic Knowledge Generation at Scale for Training Context-Augmented Multimodal LLMs

EditLord: Learning Code Transformation Rules for Code Editing

Pairwise Maximum Likelihood For Multi-Class Logistic Regression Model With Multiple Rare Classes

Optimal Auction Design in the Joint Advertising

TIMING: Temporality-Aware Integrated Gradients for Time Series Explanation

Temporal Query Network for Efficient Multivariate Time Series Forecasting

VerbalTS: Generating Time Series from Texts

Breaking the Curse of Multiagency in Robust Multi-Agent Reinforcement Learning

LIVS: A Pluralistic Alignment Dataset for Inclusive Public Spaces

AI for Global Climate Cooperation: Modeling Global Climate Negotiations, Agreements, and Long-Term Cooperation in RICE-N

Sparse Autoencoders for Hypothesis Generation

Structured Preconditioners in Adaptive Optimization: A Unified Analysis

Armijo Line-search Can Make (Stochastic) Gradient Descent Provably Faster

MERIT: Maximum-normalized Element-wise Ratio for Language Model Large-batch Training

Tilted Sharpness-Aware Minimization

Enforcing Idempotency in Neural Networks

Nonlinearly Preconditioned Gradient Methods under Generalized Smoothness

Linear convergence of Sinkhorn's algorithm for generalized static Schrödinger bridge

Provable and Practical Online Learning Rate Adaptation with Hypergradient Descent

On Linear Convergence in Smooth Convex-Concave Bilinearly-Coupled Saddle-Point Optimization: Lower Bounds and Optimal Algorithms

CaDA: Cross-Problem Routing Solver with Constraint-Aware Dual-Attention

Discrete and Continuous Difference of Submodular Minimization

Self-Supervised Transformers as Iterative Solution Improvers for Constraint Satisfaction

GPTAQ: Efficient Finetuning-Free Quantization for Asymmetric Calibration

DSBRouter: End-to-end Global Routing via Diffusion Schr\"{o}dinger Bridge

Regularized Langevin Dynamics for Combinatorial Optimization

EquivaMap: Leveraging LLMs for Automatic Equivalence Checking of Optimization Formulations

Scalable First-order Method for Certifying Optimal k-Sparse GLMs

FSL-SAGE: Accelerating Federated Split Learning via Smashed Activation Gradient Estimation

HALoS: Hierarchical Asynchronous Local SGD over Slow Networks for Geo-Distributed Large Language Model Training

Achieving Linear Speedup and Near-Optimal Complexity for Decentralized Optimization over Row-stochastic Networks

Sample, Scrutinize and Scale: Effective Inference-Time Search by Scaling Verification

Efficiently Serving Large Multimodal Models Using EPD Disaggregation

Behavior-Regularized Diffusion Policy Optimization for Offline Reinforcement Learning

Learning to Trust Bellman Updates: Selective State-Adaptive Regularization for Offline RL

Stable Offline Value Function Learning with Bisimulation-based Representations

Decision Mixer: Integrating Long-term and Local Dependencies via Dynamic Token Selection for Decision-Making

Latent Action Learning Requires Supervision in the Presence of Distractors

Accurate and Efficient World Modeling with Masked Latent Transformers

Safety-Polarized and Prioritized Reinforcement Learning

MENTOR: Mixture-of-Experts Network with Task-Oriented Perturbation for Visual Reinforcement Learning

Continual Reinforcement Learning by Planning with Online World Models

Active Fine-Tuning of Multi-Task Policies

Gradient Boosting Reinforcement Learning

Robust Offline Reinforcement Learning with Linearly Structured $f$-Divergence Regularization

When Maximum Entropy Misleads Policy Optimization

NBDI: A Simple and Effective Termination Condition for Skill Extraction from Task-Agnostic Demonstrations

Inverse Optimization via Learning Feasible Regions

Fast, Accurate Manifold Denoising by Tunneling Riemannian Optimization

Variance-Reduced Forward-Reflected-Backward Splitting Methods for Nonmonotone Generalized Equations

Non-Asymptotic and Non-Lipschitzian Bounds on Optimal Values in Stochastic Optimization Under Heavy Tails

Stacey: Promoting Stochastic Steepest Descent via Accelerated $\ell_p$-Smooth Nonconvex Optimization

Quantum Optimization via Gradient-Based Hamiltonian Descent

Contextual Optimization Under Model Misspecification: A Tractable and Generalizable Approach

Enhancing Parallelism in Decentralized Stochastic Convex Optimization

RobustZero: Enhancing MuZero Reinforcement Learning Robustness to State Perturbations

Knowledge Retention in Continual Model-Based Reinforcement Learning

Mitigating Plasticity Loss in Continual Reinforcement Learning by Reducing Churn

Value-Based Deep RL Scales Predictably

LipsNet++: Unifying Filter and Controller into a Policy Network

ARS: Adaptive Reward Scaling for Multi-Task Reinforcement Learning

SOLD: Slot Object-Centric Latent Dynamics Models for Relational Manipulation Learning from Pixels

SENSEI: Semantic Exploration Guided by Foundation Models to Learn Versatile World Models

Return Capping: Sample Efficient CVaR Policy Gradient Optimisation

Imitation Learning from a Single Temporally Misaligned Video

Explicit Exploration for High-Welfare Equilibria in Game-Theoretic Multiagent Reinforcement Learning

Risk-Sensitive Theory of Mind: Coordinating with Agents of Unknown Bias using Cumulative Prospect Theory

Finite-Sample Convergence Bounds for Trust Region Policy Optimization in Mean Field Games

Leveraging Skills from Unlabeled Prior Data for Efficient Online Exploration

Novelty Detection in Reinforcement Learning with World Models

Local Manifold Approximation and Projection for Manifold-Aware Diffusion Planning

Monte Carlo Tree Diffusion for System 2 Planning

Provable Policy Gradient for Robust Average-Reward MDPs Beyond Rectangularity

Offline Opponent Modeling with Truncated Q-driven Instant Policy Refinement

Fundamental Bias in Inverting Random Sampling Matrices with Application to Sub-sampled Newton

Theoretical guarantees on the best-of-n alignment policy

Convergence of Consistency Model with Multistep Sampling under General Data Assumptions

Revisiting the Predictability of Performative, Social Events

Fundamental limits of learning in sequence multi-index models and deep attention networks: high-dimensional asymptotics and sharp thresholds

PDUDT: Provable Decentralized Unlearning under Dynamic Topologies

Laplace Transform Based Low-Complexity Learning of Continuous Markov Semigroups

Learning Parametric Distributions from Samples and Preferences

PAC Learning with Improvements

Distributed Nonparametric Estimation: from Sparse to Dense Samples per Terminal

Metastable Dynamics of Chain-of-Thought Reasoning: Provable Benefits of Search, RL and Distillation

An Entropy-Based Model for Hierarchical Learning

Heterogeneous Data Game: Characterizing the Model Competition Across Multiple Data Sources

Provably Efficient Algorithm for Best Scoring Rule Identification in Online Principal-Agent Information Acquisition

Learning to Steer Learners in Games

Maintaining Proportional Committees with Dynamic Candidate Sets

Selective Response Strategies for GenAI

A Market for Accuracy: Classification Under Competition

Collaborative Mean Estimation Among Heterogeneous Strategic Agents: Individual Rationality, Fairness, and Truthful Contribution

Heterogeneous Label Shift: Theory and Algorithm

Towards a Formal Theory of Representational Compositionality

Test-Time Training Provably Improves Transformers as In-context Learners

Can Diffusion Models Learn Hidden Inter-Feature Rules Behind Images?

Can Transformers Reason Logically? A Study in SAT Solving

Probabilistic Factorial Experimental Design for Combinatorial Interventions

Nonconvex Theory of $M$-estimators with Decomposable Regularizers

Improved Learning via k-DTW: A Novel Dissimilarity Measure for Curves

Training Dynamics of In-Context Learning in Linear Attention

Certifiably Robust Model Evaluation in Federated Learning under Meta-Distributional Shifts

Unified Analysis of Continuous Weak Features Learning with Applications to Learning from Missing Data

A Generalization Theory for Zero-Shot Prediction

Limitations of measure-first protocols in quantum machine learning

Dimension-Independent Rates for Structured Neural Density Estimation

The Power of Random Features and the Limits of Distribution-Free Gradient Descent

ATA: Adaptive Task Allocation for Efficient Resource Management in Distributed Machine Learning

Learning to Incentivize in Repeated Principal-Agent Problems with Adversarial Agent Arrivals

Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity

Fixing the Loose Brake: Exponential-Tailed Stopping Time in Best Arm Identification

Zero-Inflated Bandits

Avoiding Catastrophe in Online Learning by Asking for Help

Implicit Riemannian Optimism with Applications to Min-Max Problems

Online Learning with Unknown Constraints

Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective

A Near Linear Query Lower Bound for Submodular Maximization

Exploiting Similarity for Computation and Communication-Efficient Decentralized Optimization

Faster Stochastic Optimization with Arbitrary Delays via Adaptive Asynchronous Mini-Batching

General framework for online-to-nonconvex conversion: Schedule-free SGD is also effective for nonconvex optimization

(ends 1:30 PM)

Poster Session 3 East [11:00-1:30]

Posters 11:00-1:30

Extractive Structures Learned in Pretraining Enable Generalization on Finetuned Facts

Internal Causal Mechanisms Robustly Predict Language Model Out-of-Distribution Behaviors

Confidential Guardian: Cryptographically Prohibiting the Abuse of Model Abstention

Pointwise Information Measures as Confidence Estimators in Deep Neural Networks: A Comparative Study

Leveraging Predictive Equivalence in Decision Trees

From Black Boxes to Transparent Minds: Evaluating and Enhancing the Theory of Mind in Multimodal Large Language Models

The Elicitation Game: Evaluating Capability Elicitation Techniques

Safety Alignment Can Be Not Superficial With Explicit Safety Signals

Policy Design for Two-sided Platforms with Participation Dynamics

Fair Clustering via Alignment

Concept-Based Unsupervised Domain Adaptation

On the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents

Correlated Errors in Large Language Models

STAMP Your Content: Proving Dataset Membership via Watermarked Rephrasings

Variational Learning of Fractional Posteriors

Censor Dependent Variational Inference

Identification of Latent Confounders via Investigating the Tensor Ranks of the Nonlinear Observations

Efficiently Vectorized MCMC on Modern Accelerators

NETS: A Non-equilibrium Transport Sampler

Progressive Tempering Sampler with Diffusion

The Noisy Laplacian: a Threshold Phenomenon for Non-Linear Dimension Reduction

Demeaned Sparse: Efficient Anomaly Detection by Residual Estimate

Interaction-Aware Gaussian Weighting for Clustered Federated Learning

Robust Conformal Outlier Detection under Contaminated Reference Data

Adjusting Model Size in Continual Gaussian Processes: How Big is Big Enough?

Can Transformers Learn Full Bayesian Inference in Context?

FAB-PPI: Frequentist, Assisted by Bayes, Prediction-Powered Inference

Linearization Turns Neural Operators into Function-Valued Gaussian Processes

Polynomial-Delay MAG Listing with Novel Locally Complete Orientation Rules

Non-asymptotic Error Bounds in $\mathcal{W}_2$-Distance with Sqrt(d) Dimension Dependence and First Order Convergence for Langevin Monte Carlo beyond Log-Concavity

CACTI: Leveraging Copy Masking and Contextual Information to Improve Tabular Data Imputation

Towards Understanding Parametric Generalized Category Discovery on Graphs

Diversified Flow Matching with Translation Identifiability

Self-Discriminative Modeling for Anomalous Graph Detection

Rethinking Confidence Scores and Thresholds in Pseudolabeling-based SSL

Point-Level Topological Representation Learning on Point Clouds

A Selective Learning Method for Temporal Graph Continual Learning

Positive-unlabeled AUC Maximization under Covariate Shift

No Task Left Behind: Isotropic Model Merging with Common and Task-Specific Subspaces

Proxy-FDA: Proxy-based Feature Distribution Alignment for Fine-tuning Vision Foundation Models without Forgetting

Interchangeable Token Embeddings for Extendable Vocabulary and Alpha-Equivalence

EnsLoss: Stochastic Calibrated Loss Ensembles for Preventing Overfitting in Classification

Understanding the Unfairness in Network Quantization

Enhancing Logits Distillation with Plug&Play Kendall's $\tau$ Ranking Loss

Principled Algorithms for Optimizing Generalized Metrics in Binary Classification

Generative Modeling Reinvents Supervised Learning: Label Repurposing with Predictive Consistency Learning

Lightweight Dataset Pruning without Full Training via Example Difficulty and Prediction Uncertainty

CAT Merging: A Training-Free Approach for Resolving Conflicts in Model Merging

Test-time Adaptation on Graphs via Adaptive Subgraph-based Selection and Regularized Prototypes

Low-Rank Tensor Transitions (LoRT) for Transferable Tensor Regression

Dual Feature Reduction for the Sparse-group Lasso and its Adaptive Variant

Computing Optimal Transport Maps and Wasserstein Barycenters Using Conditional Normalizing Flows

Near-optimal Sketchy Natural Gradients for Physics-Informed Neural Networks

ELMO : Efficiency via Low-precision and Peak Memory Optimization in Large Output Spaces

Permutation-Free High-Order Interaction Tests

Spherical Rotation Dimension Reduction with Geometric Loss Functions

Emergence in non-neural models: grokking modular arithmetic via average gradient outer product

Towards the Causal Complete Cause of Multi-Modal Representation Learning

SHE: Streaming-media Hashing Retrieval

Does Data Scaling Lead to Visual Compositional Generalization?

Active Learning with Selective Time-Step Acquisition for PDEs

FisherSFT: Data-Efficient Supervised Fine-Tuning of Language Models Using Information Gain

The Relationship Between No-Regret Learning and Online Conformal Prediction

SEAD: Unsupervised Ensemble of Streaming Anomaly Detectors

Learning Vision and Language Concepts for Controllable Image Generation

Revisiting Neural Networks for Few-Shot Learning: A Zero-Cost NAS Perspective

Sanity Checking Causal Representation Learning on a Simple Real-World System

Deep Fuzzy Multi-view Learning for Reliable Classification

DANCE: Dual Unbiased Expansion with Group-acquired Alignment for Out-of-distribution Graph Fairness Learning

PlaySlot: Learning Inverse Latent Dynamics for Controllable Object-Centric Video Prediction and Planning

Tensor Decomposition Based Memory-Efficient Incremental Learning

Adaptive Data Collection for Robust Learning Across Multiple Distributions

Unconstrained Robust Online Convex Optimization

QUTE: Quantifying Uncertainty in TinyML models with Early-exit-assisted ensembles for model-monitoring

Accelerating Linear Recurrent Neural Networks for the Edge with Unstructured Sparsity

Compute or Load KV Cache? Why Not Both?

KoopSTD: Reliable Similarity Analysis between Dynamical Systems via Approximating Koopman Spectrum with Timescale Decoupling

RAGGED: Towards Informed Design of Scalable and Stable RAG Systems

Prompt-to-Leaderboard: Prompt-Adaptive LLM Evaluations

am-ELO: A Stable Framework for Arena-based LLM Evaluation

Tightening Causal Bounds via Covariate-Aware Optimal Transport

Integer Programming for Generalized Causal Bootstrap Designs

Automatic Reward Shaping from Confounded Offline Data

Identifying Causal Direction via Variational Bayesian Compression

Doubly Robust Fusion of Many Treatments for Policy Learning

From Logits to Hierarchies: Hierarchical Clustering made Simple

MCU: An Evaluation Framework for Open-Ended Game Agents

From Jack of All Trades to Master of One: Specializing LLM-based Autoraters to a Test Set

David and Goliath: Small One-step Model Beats Large Diffusion with Score Post-training

AutoEval Done Right: Using Synthetic Data for Model Evaluation

RE-IMAGINE: Symbolic Benchmark Synthesis for Reasoning Evaluation

Bridging Fairness and Efficiency in Conformal Inference: A Surrogate-Assisted Group-Clustered Approach

Overestimation in LLM Evaluation: A Controlled Large-Scale Study on Data Contamination’s Impact on Machine Translation

Permutation-based Rank Test in the Presence of Discretization and Application in Causal Discovery with Mixed Data

Double Machine Learning for Causal Inference under Shared-State Interference

Stronger Neyman Regret Guarantees for Adaptive Experimental Design

On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization

An Improved Clique-Picking Algorithm for Counting Markov Equivalent DAGs via Super Cliques Transfer

Multi-Objective Causal Bayesian Optimization

MetricEmbedding: Accelerate Metric Nearness by Tropical Inner Product

Best Subset Selection: Optimal Pursuit for Feature Selection and Elimination

Direct Prediction Set Minimization via Bilevel Conformal Classifier Training

COExpander: Adaptive Solution Expansion for Combinatorial Optimization

Distributional Diffusion Models with Scoring Rules

One-Step Generalization Ratio Guided Optimization for Domain Generalization

Foundation Model Insights and a Multi-Model Approach for Superior Fine-Grained One-shot Subset Selection

Learning In-context $n$-grams with Transformers: Sub-$n$-grams Are Near-Stationary Points

Hide & Seek: Transformer Symmetries Obscure Sharpness & Riemannian Geometry Finds It

A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models

On the Impact of Hard Adversarial Instances on Overfitting in Adversarial Training

Outlier Gradient Analysis: Efficiently Identifying Detrimental Training Samples for Deep Learning Models

Function-Space Learning Rates

Towards Memorization Estimation: Fast, Formal and Free

Lego Sketch: A Scalable Memory-augmented Neural Network for Sketching Data Streams

Knowledge Swapping via Learning and Unlearning

On the Power of Learning-Augmented Search Trees

Efficient Heterogeneity-Aware Federated Active Data Selection

Optimal transport-based conformal prediction

Multi-Session Budget Optimization for Forward Auction-based Federated Learning

BCE vs. CE in Deep Feature Learning

Compositional Risk Minimization

On the Provable Separation of Scales in Maximal Update Parameterization

Modular Duality in Deep Learning

On the Clean Generalization and Robust Overfitting in Adversarial Training from Two Theoretical Views: Representation Complexity and Training Dynamics

TransPL: VQ-Code Transition Matrices for Pseudo-Labeling of Time Series Unsupervised Domain Adaptation

Understanding and Improving Length Generalization in Recurrent Models

LETS Forecast: Learning Embedology for Time Series Forecasting

LAST SToP for Modeling Asynchronous Time Series

LangTime: A Language-Guided Unified Model for Time Series Forecasting with Proximal Policy Optimization

LightGTS: A Lightweight General Time Series Forecasting Model

No Soundness in the Real World: On the Challenges of the Verification of Deployed Neural Networks

Pixel2Feature Attack (P2FA): Rethinking the Perturbed Space to Enhance Adversarial Transferability

Backdoor Attacks in Token Selection of Attention Mechanism

ERICT: Enhancing Robustness by Identifying Concept Tokens in Zero-Shot Vision Language Models

LEVIS: Large Exact Verifiable Input Spaces for Neural Networks

Learning State-Based Node Representations from a Class Hierarchy for Fine-Grained Open-Set Detection

IT$^3$: Idempotent Test-Time Training

Sample-specific Noise Injection for Diffusion-based Adversarial Purification

SDP-CROWN: Efficient Bound Propagation for Neural Network Verification with Tightness of Semidefinite Programming

One Leaf Reveals the Season: Occlusion-Based Contrastive Learning with Semantic-Aware Views for Efficient Visual Representation

GraphCL: Graph-based Clustering for Semi-Supervised Medical Image Segmentation

ExPLoRA: Parameter-Efficient Extended Pre-Training to Adapt Vision Transformers under Domain Shifts

Training High Performance Spiking Neural Network by Temporal Model Calibration

ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks

NeuronTune: Towards Self-Guided Spurious Bias Mitigation

CoPINN: Cognitive Physics-Informed Neural Networks

Approximating Latent Manifolds in Neural Networks via Vanishing Ideals

GTR: A General, Multi-View, and Dynamic Framework for Trajectory Representation Learning

A Chaotic Dynamics Framework Inspired by Dorsal Stream for Event Signal Processing

MoEQuant: Enhancing Quantization for Mixture-of-Experts Large Language Models via Expert-Balanced Sampling and Affinity Guidance

DLP: Dynamic Layerwise Pruning in Large Language Models

Liger: Linearizing Large Language Models to Gated Recurrent Structures

Discriminative Finetuning of Generative Large Language Models without Reward Models and Human Preference Data

Distillation Scaling Laws

GuardAgent: Safeguard LLM Agents via Knowledge-Enabled Reasoning

NICE Data Selection for Instruction Tuning in LLMs with Non-differentiable Evaluation Metric

Emergent Response Planning in LLMs

Forest-of-Thought: Scaling Test-Time Compute for Enhancing LLM Reasoning

Reward-Guided Speculative Decoding for Efficient LLM Reasoning

A Unified Approach to Routing and Cascading for LLMs

Self-Consistency Preference Optimization

Free Process Rewards without Process Labels

Determining Layer-wise Sparsity for Large Language Models Through a Theoretical Perspective

rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking

De-mark: Watermark Removal in Large Language Models

ResQ: Mixed-Precision Quantization of Large Language Models with Low-Rank Residuals

Inference-Time Decomposition of Activations (ITDA): A Scalable Approach to Interpreting Large Language Models

DyCodeEval: Dynamic Benchmarking of Reasoning Capabilities in Code Large Language Models Under Data Contamination

Fast Video Generation with Sliding Tile Attention

Evaluating Judges as Evaluators: The JETTS Benchmark of LLM-as-Judges as Test-Time Scaling Evaluators

Flow-of-Options: Diversified and Improved LLM Reasoning by Thinking Through Options

Learning Safety Constraints for Large Language Models

Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models

CoreMatching: A Co-adaptive Sparse Inference Framework with Token and Neuron Pruning for Comprehensive Acceleration of Vision-Language Models

Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction

Why Has Predicting Downstream Capabilities of Frontier AI Models with Scale Remained Elusive?

FlipAttack: Jailbreak LLMs via Flipping

Mastering Board Games by External and Internal Planning with Language Models

AlphaPO: Reward Shape Matters for LLM Alignment

MAPLE: Many-Shot Adaptive Pseudo-Labeling for In-Context Learning

Guided Search Strategies in Non-Serializable Environments with Applications to Software Engineering Agents

MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations

Organize the Web: Constructing Domains Enhances Pre-Training Data Curation

MARGE: Improving Math Reasoning with Guided Exploration

ConfPO: Exploiting Policy Model Confidence for Critical Token Selection in Preference Optimization

BlockDialect: Block-wise Fine-grained Mixed Format Quantization for Energy-Efficient LLM Inference

Communicating Activations Between Language Model Agents

AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism

RWKVQuant: Quantizing the RWKV Family with Proxy Guided Hybrid of Scalar and Vector Quantization

The Emperor's New Clothes in Benchmarking? A Rigorous Examination of Mitigation Strategies for LLM Benchmark Data Contamination

On the Robustness of Reward Models for Language Model Alignment

Federated In-Context Learning: Iterative Refinement for Improved Answer Quality

Improving Diversity in Language Models: When Temperature Fails, Change the Loss

NExtLong: Toward Effective Long-Context Training without Long Documents

SliM-LLM: Salience-Driven Mixed-Precision Quantization for Large Language Models

Right Now, Wrong Then: Non-Stationary Direct Preference Optimization under Preference Drift

Bounded Rationality for LLMs: Satisficing Alignment at Inference-Time

LLMs can see and hear without any training

Investigating Non-Transitivity in LLM-as-a-Judge

Aligning LLMs by Predicting Preferences from User Writing Samples

Look Twice Before You Answer: Memory-Space Visual Retracing for Hallucination Mitigation in Multimodal Large Language Models

RoSTE: An Efficient Quantization-Aware Supervised Fine-Tuning Approach for Large Language Models

Understanding Multimodal LLMs Under Distribution Shifts: An Information-Theoretic Approach

Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection

Explicit Preference Optimization: No Need for an Implicit Reward Model

Radio: Rate–Distortion Optimization for Large Language Model Compression

Instruction-Following Pruning for Large Language Models

Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areas

Automated Benchmark Generation for Repository-Level Coding Tasks

µnit Scaling: Simple and Scalable FP8 LLM Training

Divide and Conquer: Grounding LLMs as Efficient Decision-Making Agents via Offline Hierarchical Reinforcement Learning

UDora: A Unified Red Teaming Framework against LLM Agents by Dynamically Hijacking Their Own Reasoning

DataDecide: How to Predict Best Pretraining Data with Small Experiments

LensLLM: Unveiling Fine-Tuning Dynamics for LLM Selection

SafeAuto: Knowledge-Enhanced Safe Autonomous Driving with Multimodal Foundation Models

Chameleon: A Flexible Data-mixing Framework for Language Model Pretraining and Finetuning

TokenSwift: Lossless Acceleration of Ultra Long Sequence Generation

Grokking in the Wild: Data Augmentation for Real-World Multi-Hop Reasoning with Transformers

Parameters vs FLOPs: Scaling Laws for Optimal Sparsity for Mixture-of-Experts Language Models

EvoPress: Accurate Dynamic Model Compression via Evolutionary Search

VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data

Principled Data Selection for Alignment: The Hidden Risks of Difficult Examples

Flow of Reasoning: Training LLMs for Divergent Reasoning with Minimal Examples

RATE: Causal Explainability of Reward Models with Imperfect Counterfactuals

ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning

BEST-Route: Adaptive LLM Routing with Test-Time Optimal Compute

Token Cleaning: Fine-Grained Data Selection for LLM Supervised Fine-Tuning

FactTest: Factuality Testing in Large Language Models with Finite-Sample and Distribution-Free Guarantees

Can MLLMs Reason in Multimodality? EMMA: An Enhanced MultiModal ReAsoning Benchmark

Are LLMs Prescient? A Continuous Evaluation using Daily News as the Oracle

MetaAgent: Automatically Constructing Multi-Agent Systems Based on Finite State Machines

Occult: Optimizing Collaborative Communications across Experts for Accelerated Parallel MoE Training and Inference

Mitigating Object Hallucination in Large Vision-Language Models via Image-Grounded Guidance

When Do LLMs Help With Node Classification? A Comprehensive Analysis

Mitigating Over-Squashing in Graph Neural Networks by Spectrum-Preserving Sparsification

Discrete Neural Algorithmic Reasoning

CEGA: A Cost-Effective Approach for Graph-Based Model Extraction and Acquisition

Uncertainty Estimation for Heterophilic Graphs Through the Lens of Information Theory

Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment

SHARP-Distill: A 68× Faster Recommender System with Hypergraph Neural Networks and Language Models

Commute Graph Neural Networks

Dynamic Mixture of Curriculum LoRA Experts for Continual Multimodal Instruction Tuning

Can Large Language Models Understand Intermediate Representations in Compilers?

Thinking LLMs: General Instruction Following with Thought Generation

video-SALMONN-o1: Reasoning-enhanced Audio-visual Large Language Model

How Contaminated Is Your Benchmark? Measuring Dataset Leakage in Large Language Models with Kernel Divergence

Improving the Effective Receptive Field of Message-Passing Neural Networks

TopoTune: A Framework for Generalized Combinatorial Complex Neural Networks

Can Classic GNNs Be Strong Baselines for Graph-level Tasks? Simple Architectures Meet Excellence

Geometric Hyena Networks for Large-scale Equivariant Learning

Cooperation of Experts: Fusing Heterogeneous Information with Large Margin

COMRECGC: Global Graph Counterfactual Explainer through Common Recourse

Outlier-Aware Post-Training Quantization for Discrete Graph Diffusion Models

What Makes a Good Feedforward Computational Graph?

GCAL: Adapting Graph Models to Evolving Domain Shifts

Improving the Scaling Laws of Synthetic Data with Deliberate Practice

Geometric Generative Modeling with Noise-Conditioned Graph Networks

AsymRnR: Video Diffusion Transformers Acceleration with Asymmetric Reduction and Restoration

Learning Extrapolative Sequence Transformations from Markov Chains

Improving Consistency Models with Generator-Augmented Flows

Ergodic Generative Flows

Does Generation Require Memorization? Creative Diffusion Models using Ambient Diffusion

Robust Multi-bit Text Watermark with LLM-based Paraphrasers

Improved Discretization Complexity Analysis of Consistency Models: Variance Exploding Forward Process and Decay Discretization Scheme

Generative Data Mining with Longtail-Guided Diffusion

Diffusion Adversarial Post-Training for One-Step Video Generation

Harmonizing Geometry and Uncertainty: Diffusion with Hyperspheres

IMPACT: Iterative Mask-based Parallel Decoding for Text-to-Audio Generation with Diffusion Modeling

Field Matching: an Electrostatic Paradigm to Generate and Transfer Data

S4S: Solving for a Fast Diffusion Model Solver

DiTAR: Diffusion Transformer Autoregressive Modeling for Speech Generation

Geometric Representation Condition Improves Equivariant Molecule Generation

EasyRef: Omni-Generalized Group Image Reference for Diffusion Models via Multimodal LLM

Morse: Dual-Sampling for Lossless Acceleration of Diffusion Models

Symmetry-Aware GFlowNets

Q-VDiT: Towards Accurate Quantization and Distillation of Video-Generation Diffusion Transformers

FlexControl: Computation-Aware Conditional Control with Differentiable Router for Text-to-Image Generation

Categorical Schrödinger Bridge Matching

Variational Rectified Flow Matching

TabSDS: a Lightweight, Fully Non-Parametric, and Model Free Approach for Generating Synthetic Tabular Data

Stochastic Forward–Backward Deconvolution: Training Diffusion Models with Finite Noisy Datasets

Controlled Generation with Equivariant Variational Flow Matching

Target Concrete Score Matching: A Holistic Framework for Discrete Diffusion

Be a Goldfish: Forgetting Bad Conditioning in Sparse Linear Regression via Variational Autoencoders

Discrete Markov Probabilistic Models: An Improved Discrete Score-Based Framework with sharp convergence bounds under minimal assumptions

Bayesian Neural Scaling Law Extrapolation with Prior-Data Fitted Networks

Make LoRA Great Again: Boosting LoRA with Adaptive Singular Values and Mixture-of-Experts Optimization Alignment

Diffusion Instruction Tuning

Modulated Diffusion: Accelerating Generative Modeling with Modulated Quantization

Continuous Semi-Implicit Models

WMAdapter: Adding WaterMark Control to Latent Diffusion Models

INRFlow: Flow Matching for INRs in Ambient Space

Efficient Fine-Grained Guidance for Diffusion Model Based Symbolic Music Generation

EasyInv: Toward Fast and Better DDIM Inversion

Inverse Problem Sampling in Latent Space Using Sequential Monte Carlo

Visual Generation Without Guidance

Spatial Reasoning with Denoising Models

Peri-LN: Revisiting Normalization Layer in the Transformer Architecture

From Low Rank Gradient Subspace Stabilization to Low-Rank Weights: Observations, Theories, and Applications

Primitive Vision: Improving Diagram Understanding in MLLMs

Attention-Level Speculation

Robust Noise Attenuation via Adaptive Pooling of Transformer Outputs

Primphormer: Efficient Graph Transformers with Primal Representations

DeepCrossAttention: Supercharging Transformer Residual Connections

Curvature-aware Graph Attention for PDEs on Manifolds

Visual Attention Never Fades: Selective Progressive Attention ReCalibration for Detailed Image Captioning in Multimodal Large Language Models

SAFE: Finding Sparse and Flat Minima to Improve Pruning

Transolver++: An Accurate Neural Solver for PDEs on Million-Scale Geometries

SWAN: SGD with Normalization and Whitening Enables Stateless LLM Training

Learning-Augmented Hierarchical Clustering

On Fine-Grained Distinct Element Estimation

Safety Certificate against Latent Variables with Partially Unidentifiable Dynamics

Learning curves theory for hierarchically compositional data with power-law distributed features

Harnessing Heterogeneous Statistical Strength for Personalized Federated Learning via Hierarchical Bayesian Inference

SEFE: Superficial and Essential Forgetting Eliminator for Multimodal Continual Instruction Tuning

Matryoshka Quantization

Prompt-based Depth Pruning of Large Language Models

Addressing Imbalanced Domain-Incremental Learning through Dual-Balance Collaborative Experts

Merge-Friendly Post-Training Quantization for Multi-Target Domain Adaptation

Learning to Stop: Deep Learning for Mean Field Optimal Stopping

Provably Improving Generalization of Few-shot models with Synthetic Data

QuEST: Stable Training of LLMs with 1-Bit Weights and Activations

The Sharpness Disparity Principle in Transformers for Accelerating Language Model Pre-Training

Position: Societal Impacts Research Requires Benchmarks for Creative Composition Tasks

Position: Retrieval-augmented systems can be dangerous medical communicators

Position: Contextual Integrity is Inadequately Applied to Language Models

Position: Evaluating Generative AI Systems Is a Social Science Measurement Challenge

Position: It Is Time We Test Neural Computation In Vitro

Position: Uncertainty Quantification Needs Reassessment for Large Language Model Agents

Position: AI Evaluation Should Learn from How We Test Humans

Do Not Mimic My Voice : Speaker Identity Unlearning for Zero-Shot Text-to-Speech

Adversarial Inception Backdoor Attacks against Reinforcement Learning

Decision Making under the Exponential Family: Distributionally Robust Optimisation with Bayesian Ambiguity Sets

Optimal Survey Design for Private Mean Estimation

Position: When Incentives Backfire, Data Stops Being Human

Position: Rethinking LLM Bias Probing Using Lessons from the Social Sciences

Position: General Intelligence Requires Reward-based Pretraining

Weak-to-Strong Jailbreaking on Large Language Models

SafeArena: Evaluating the Safety of Autonomous Web Agents

X-Transfer Attacks: Towards Super Transferable Adversarial Attacks on CLIP

Emoji Attack: Enhancing Jailbreak Attacks Against Judge LLM Detection

Mind the Gap: A Practical Attack on GGUF Quantization

Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning

SafetyAnalyst: Interpretable, Transparent, and Steerable Safety Moderation for AI Behavior

LEMoN: Label Error Detection using Multimodal Neighbors

EraseAnything: Enabling Concept Erasure in Rectified Flow Transformers

POROver: Improving Safety and Reducing Overrefusal in Large Language Models with Overgeneration and Preference Optimization

$S^2$FGL: Spatial Spectral Federated Graph Learning

Minimalist Concept Erasure in Generative Models

Adversarial Reasoning at Jailbreaking Time

LlavaGuard: An Open VLM-based Framework for Safeguarding Vision Datasets and Models

Differentially Private Analysis for Binary Response Models: Optimality, Estimation, and Inference

DMM: Distributed Matrix Mechanism for Differentially-Private Federated Learning Based on Constant-Overhead Linear Secret Resharing

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

PASS: Private Attributes Protection with Stochastic Data Substitution

Plausible Token Amplification for Improving Accuracy of Differentially Private In-Context Learning Based on Implicit Bayesian Inference

Is Your Model Fairly Certain? Uncertainty-Aware Fairness Evaluation for LLMs

On the Alignment between Fairness and Accuracy: from the Perspective of Adversarial Robustness

(ends 1:30 PM)

1 p.m.

Break:

Hosted Grab and Go Lunch

(ends 2:00 PM)

2 p.m.

Invited Talk:

Adaptive Alignment: Designing AI for a Changing World - Frauke Kreuter

Frauke Kreuter

(ends 3:00 PM)

3:30 p.m.

Oral 4A Representations 2 [3:30-4:30]

Orals 3:30-4:30

[3:30] Sundial: A Family of Highly Capable Time Series Foundation Models

[3:45] Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation

[4:00] Partition First, Embed Later: Laplacian-Based Feature Partitioning for Refined Embedding and Visualization of High-Dimensional Data

[4:15] Equivalence is All: A Unified View for Self-supervised Graph Learning

(ends 4:30 PM)

Oral 4B Positions: Generative AI Evaluation [3:30-4:30]

Orals 3:30-4:30

[3:30] Position: AI Competitions Provide the Gold Standard for Empirical Rigor in GenAI Evaluation

[3:45] Position: Medical Large Language Model Benchmarks Should Prioritize Construct Validity

[4:00] Position: Principles of Animal Cognition to Improve LLM Evaluations

[4:15] Position: Political Neutrality in AI Is Impossible — But Here Is How to Approximate It

(ends 4:30 PM)

Oral 4C Privacy and Uncertainty Quantification [3:30-4:30]

Orals 3:30-4:30

[3:30] On Differential Privacy for Adaptively Solving Search Problems via Sketching

[3:45] Going Deeper into Locally Differentially Private Graph Neural Networks

[4:00] Auditing $f$-differential privacy in one run

[4:15] Conformal Prediction as Bayesian Quadrature

(ends 4:30 PM)

Oral 4D Applications in Science and Language [3:30-4:30]

Orals 3:30-4:30

[3:30] AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models

[3:45] Long-Form Speech Generation with Spoken Language Models

[4:00] Learning Time-Varying Multi-Region Brain Communications via Scalable Markovian Gaussian Processes

[4:15] Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction

(ends 4:30 PM)

Oral 4E Algorithms [3:30-4:30]

Orals 3:30-4:30

[3:30] Improved Regret Analysis in Gaussian Process Bandits: Optimality for Noiseless Reward, RKHS norm, and Non-Stationary Variance

[3:45] High-Dimensional Prediction for Sequential Decision Making

[4:00] Near-Optimal Decision Trees in a SPLIT Second

[4:15] Expected Variational Inequalities

(ends 4:30 PM)

4:30 p.m.

Break:

Coffee Break

(ends 5:30 PM)

Poster Session 4 West [4:30-7:00]

Posters 4:30-7:00

Windows Agent Arena: Evaluating Multi-Modal OS Agents at Scale

Dimensionality Reduction on Complex Vector Spaces for Euclidean Distance with Dynamic Weights

CursorCore: Assist Programming through Aligning Anything

Fully Dynamic Embedding into $\ell_p$ Spaces

Understanding the Kronecker Matrix-Vector Complexity of Linear Algebra

Square$\chi$PO: Differentially Private and Robust $\chi^2$-Preference Optimization in Offline Direct Alignment

Accelerating Quantum Reinforcement Learning with a Quantum Natural Policy Gradient Based Approach

Bellman Unbiasedness: Toward Provably Efficient Distributional Reinforcement Learning with General Value Function Approximation

Quantum Speedups in Regret Analysis of Infinite Horizon Average-Reward Markov Decision Processes

Rectifying Conformity Scores for Better Conditional Coverage

On Differential Privacy for Adaptively Solving Search Problems via Sketching

Matrix Completion with Incomplete Side Information via Orthogonal Complement Projection

$\mathcal{V}ista\mathcal{DPO}$: Video Hierarchical Spatial-Temporal Direct Preference Optimization for Large Video Models

Generalization Principles for Inference over Text-Attributed Graphs with Large Language Models

Expected Variational Inequalities

Invariant Deep Uplift Modeling for Incentive Assignment in Online Marketing via Probability of Necessity and Sufficiency

AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML

AffectGPT: A New Dataset, Model, and Benchmark for Emotion Understanding with Multimodal Large Language Models

Skip the Equations: Learning Behavior of Personalized Dynamical Systems Directly From Data

Thermalizer: Stable autoregressive neural emulation of spatiotemporal chaos

Differential Privacy Under Class Imbalance: Methods and Empirical Insights

Sidechain conditioning and modeling for full-atom protein sequence design with FAMPNN

Foundation Molecular Grammar: Multi-Modal Foundation Models Induce Interpretable Molecular Graph Languages

Retrieval Augmented Zero-Shot Enzyme Generation for Specified Substrate

LLM-Augmented Chemical Synthesis and Design Decision Programs

Learn Singularly Perturbed Solutions via Homotopy Dynamics

Boosting Protein Graph Representations through Static-Dynamic Fusion

Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction

Retrieval-Augmented Language Model for Knowledge-aware Protein Encoding

A Physics-Augmented Deep Learning Framework for Classifying Single Molecule Force Spectroscopy Data

OmniArch: Building Foundation Model for Scientific Computing

Sub-Sequential Physics-Informed Learning with State Space Model

UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules

Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions

Latent Mamba Operator for Partial Differential Equations

Piloting Structure-Based Drug Design via Modality-Specific Optimal Schedule

CoCoA-Mix: Confusion-and-Confidence-Aware Mixture Model for Context Optimization

Attributes Shape the Embedding Space of Face Recognition Models

Agent Workflow Memory

Scaling Large Motion Models with Million-Level Human Motions

Componential Prompt-Knowledge Alignment for Domain Incremental Learning

Task-Gated Multi-Expert Collaboration Network for Degraded Multi-Modal Image Fusion

Diff-MoE: Diffusion Transformer with Time-Aware and Space-Adaptive Experts

3D Question Answering via only 2D Vision-Language Models

ReFrame: Layer Caching for Accelerated Inference in Real-Time Rendering

Beyond Cropped Regions: New Benchmark and Corresponding Baseline for Chinese Scene Text Retrieval in Diverse Layouts

UncertainSAM: Fast and Efficient Uncertainty Quantification of the Segment Anything Model

MuLan: Adapting Multilingual Diffusion Models for Hundreds of Languages with Negligible Cost

Text-to-CAD Generation Through Infusing Visual Feedback in Large Language Models

Visual Abstraction: A Plug-and-Play Approach for Text-Visual Retrieval

Visual Graph Arena: Evaluating Visual Conceptualization of Vision and Multimodal Large Language Models

Steerable Transformers for Volumetric Data

MoMa: Modulating Mamba for Adapting Image Foundation Models to Video Recognition

Event-Customized Image Generation

Probabilistic Interactive 3D Segmentation with Hierarchical Neural Processes

Unifying Knowledge from Diverse Datasets to Enhance Spatial-Temporal Modeling: A Granularity-Adaptive Geographical Embedding Approach

Inverse problems with experiment-guided AlphaFold

Rethink GraphODE Generalization within Coupled Dynamical System

SlimLLM: Accurate Structured Pruning for Large Language Models

Efficient and Separate Authentication Image Steganography Network

Subobject-level Image Tokenization

Complex Wavelet Mutual Information Loss: A Multi-Scale Loss Function for Semantic Segmentation

Gamma Distribution PCA-Enhanced Feature Learning for Angle-Robust SAR Target Recognition

Beyond the Permutation Symmetry of Transformers: The Role of Rotation for Model Fusion

AGAV-Rater: Adapting Large Multimodal Model for AI-Generated Audio-Visual Quality Assessment

Orient Anything: Learning Robust Object Orientation Estimation from Rendering 3D Models

LieRE: Lie Rotational Positional Encodings

DAMA: Data- and Model-aware Alignment of Multi-modal LLMs

Auto-reconfiguration for Latency Minimization in CPU-based DNN Serving

Cross-Modal Alignment via Variational Copula Modelling

SAFER: A Calibrated Risk-Aware Multimodal Recommendation Model for Dynamic Treatment Regimes

Distribution-aware Fairness Learning in Medical Image Segmentation From A Control-Theoretic Perspective

EARTH: Epidemiology-Aware Neural ODE with Continuous Disease Transmission Graph

CFP-Gen: Combinatorial Functional Protein Generation via Diffusion Language Models

PepTune: De Novo Generation of Therapeutic Peptides with Multi-Objective-Guided Discrete Diffusion

Protriever: End-to-End Differentiable Protein Homology Search for Fitness Prediction

Reconstructing Cell Lineage Trees from Phenotypic Features with Metric Learning

LangDAug: Langevin Data Augmentation for Multi-Source Domain Generalization in Medical Image Segmentation

Gradient-based Explanations for Deep Learning Survival Models

Breaking the Barrier of Hard Samples: A Data-Centric Approach to Synthetic Data for Medical Tasks

Closed-Loop Long-Horizon Robotic Planning via Equilibrium Sequence Modeling

Efficient Robotic Policy Learning via Latent Space Backward Planning

Learning Time-Varying Multi-Region Brain Communications via Scalable Markovian Gaussian Processes

Dynamical Modeling of Behaviorally Relevant Spatiotemporal Patterns in Neural Imaging Data

Scaling Laws for Task-Optimized Models of the Primate Visual Ventral Stream

BSO: Binary Spiking Online Optimization Algorithm

SynEVO: A neuro-inspired spatiotemporal evolutional framework for cross-domain adaptation

Promoting Ensemble Diversity with Interactive Bayesian Distributional Robustness for Fine-tuning Foundation Models

Efficient First-Order Optimization on the Pareto Set for Multi-Objective Learning under Preference Guidance

High-Fidelity Simultaneous Speech-To-Speech Translation

Adaptive Elicitation of Latent Information Using Natural Language

A Variational Framework for Improving Naturalness in Generative Spoken Language Models

Synthesizing Privacy-Preserving Text Data via Finetuning *without* Finetuning Billion-Scale LLMs

Emotional Face-to-Speech

CoSER: Coordinating LLM-Based Persona Simulation of Established Roles

Learning from others' mistakes: Finetuning machine translation models with span-level error annotations

Sortformer: A Novel Approach for Permutation-Resolved Speaker Supervision in Speech-to-Text Systems

Long-Form Speech Generation with Spoken Language Models

From Mechanistic Interpretability to Mechanistic Biology: Training, Evaluating, and Interpreting Sparse Autoencoders on Protein Language Models

Boosting Masked ECG-Text Auto-Encoders as Discriminative Learners

Domain-Adapted Diffusion Model for PROTAC Linker Design Through the Lens of Density Ratio in Chemical Space

SUICA: Learning Super-high Dimensional Sparse Implicit Neural Representations for Spatial Transcriptomics

DriveGPT: Scaling Autoregressive Behavior Models for Driving

A Physics-Informed Machine Learning Framework for Safe and Optimal Control of Autonomous Systems

DiffusionVLA: Scaling Robot Foundation Models via Unified Diffusion and Autoregression

SAM2Act: Integrating Visual Foundation Model with A Memory Architecture for Robotic Manipulation

Optimizing Social Network Interventions via Hypergradient-Based Recommender System Design

Causality-Aware Contrastive Learning for Robust Multivariate Time-Series Anomaly Detection

Exploring Representations and Interventions in Time Series Foundation Models

TimePro: Efficient Multivariate Long-term Time Series Forecasting with Variable- and Time-Aware Hyper-state

BRIDGE: Bootstrapping Text to Control Time-Series Generation via Multi-Agent Iterative Optimization and Diffusion Modeling

Latent Variable Estimation in Bayesian Black-Litterman Models

LineFlow: A Framework to Learn Active Control of Production Lines

COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning

DiMa: Understanding the Hardness of Online Matching Problems via Diffusion Models

Chip Placement with Diffusion Models

PARQ: Piecewise-Affine Regularized Quantization

Aligned Multi Objective Optimization

Towards Understanding Gradient Dynamics of the Sliced-Wasserstein Distance via Critical Point Analysis

Fast Tensor Completion via Approximate Richardson Iteration

Toward a Unified Theory of Gradient Descent under Generalized Smoothness

Online Pre-Training for Offline-to-Online Reinforcement Learning

Enhancing Diversity In Parallel Agents: A Maximum State Entropy Exploration Story

SERENA: A Unified Stochastic Recursive Variance Reduced Gradient Framework for Riemannian Non-Convex Optimization

EARL-BO: Reinforcement Learning for Multi-Step Lookahead, High-Dimensional Bayesian Optimization

Refining Adaptive Zeroth-Order Optimization at Ease

BARK: A Fully Bayesian Tree Kernel for Black-box Optimization

Provable Benefit of Random Permutations over Uniform Sampling in Stochastic Coordinate Descent

A Comprehensive Framework for Analyzing the Convergence of Adam: Bridging the Gap with SGD

The Global Convergence Time of Stochastic Gradient Descent in Non-Convex Landscapes: Sharp Estimates via Large Deviations

Stochastic Smoothed Primal-Dual Algorithms for Nonconvex Optimization with Linear Inequality Constraints

Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery

When, Where and Why to Average Weights?

Improving Generalization in Federated Learning with Highly Heterogeneous Data via Momentum-Based Stochastic Controlled Weight Averaging

Shortcut-connected Expert Parallelism for Accelerating Mixture of Experts

Nesterov Method for Asynchronous Pipeline Parallel Optimization

Breaking Barriers: Combinatorial Algorithms for Non-Monotone Submodular Maximization with Sublinear Adaptivity and $1/e$ Approximation

Edge-Colored Clustering in Hypergraphs: Beyond Minimizing Unsatisfied Edges

Faster Global Minimum Cut with Predictions

Geometric Algebra Planes: Convex Implicit Neural Volumes

The Sparse-Plus-Low-Rank Quasi-Newton Method for Entropic-Regularized Optimal Transport

Scaffold with Stochastic Gradients: New Analysis with Linear Speed-Up

Maximum Total Correlation Reinforcement Learning

Penalizing Infeasible Actions and Reward Scaling in Reinforcement Learning with Offline Data

Policy Regularization on Globally Accessible States in Cross-Dynamics Reinforcement Learning

Online Laplacian-Based Representation Learning in Reinforcement Learning

Meta-Reinforcement Learning with Adaptation from Human Feedback via Preference-Order-Preserving Task Embedding

Robot-Gated Interactive Imitation Learning with Adaptive Intervention Mechanism

Score as Action: Fine Tuning Diffusion Generative Models by Continuous-time Reinforcement Learning

Consensus Based Stochastic Optimal Control

RobustLight: Improving Robustness via Diffusion Reinforcement Learning for Traffic Signal Control

LMRL Gym: Benchmarks for Multi-Turn Reinforcement Learning with Language Models

Proto Successor Measure: Representing the Behavior Space of an RL Agent

CLARIFY: Contrastive Preference Reinforcement Learning for Untangling Ambiguous Queries

A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks

Flow Q-Learning

Mastering Massive Multi-Task Reinforcement Learning via Mixture-of-Expert Decision Transformer

Hierarchical Reinforcement Learning with Uncertainty-Guided Diffusional Subgoals

Efficient Skill Discovery via Regret-Aware Optimization

DIME: Diffusion-Based Maximum Entropy Reinforcement Learning

Comparing Comparisons: Informative and Easy Human Feedback with Distinguishability Queries

The Courage to Stop: Overcoming Sunk Cost Fallacy in Deep Reinforcement Learning

Generalization Analysis for Controllable Learning

Graph Attention is Not Always Beneficial: A Theoretical Analysis of Graph Attention Mechanisms via Contextual Stochastic Block Models

Mastering Multiple-Expert Routing: Realizable $H$-Consistency and Strong Guarantees for Learning to Defer

Learning Configurations for Data-Driven Multi-Objective Optimization

Anytime-Constrained Equilibria in Polynomial Time

Continuous-Time Analysis of Heavy Ball Momentum in Min-Max Games

A Closer Look at Multimodal Representation Collapse

A Theoretical Study of (Hyper) Self-Attention through the Lens of Interactions: Representation, Training, Generalization

Improved Expressivity of Hypergraph Neural Networks through High-Dimensional Generalized Weisfeiler-Leman Algorithms

Propagation of Chaos for Mean-Field Langevin Dynamics and its Application to Model Ensemble

Universal Length Generalization with Turing Programs

Grokking Beyond the Euclidean Norm of Model Parameters

Sharp Optimality of Simple, Plug-in Estimation of the Fisher Information of a Smoothed Density

Towards Better-than-2 Approximation for Constrained Correlation Clustering

Empirical Design in Reinforcement Learning

Improving the Continuity of Goal-Achievement Ability via Policy Self-Regularization for Goal-Conditioned Reinforcement Learning

Power Mean Estimation in Stochastic Continuous Monte-Carlo Tree Search

Finite-Time Global Optimality Convergence in Deep Neural Actor-Critic Methods for Decentralized Multi-Agent Reinforcement Learning

Sable: a Performant, Efficient and Scalable Sequence Model for MARL

Zero Shot Generalization of Vision-Based RL Without Data Augmentation

Efficient Multivariate Robust Mean Estimation Under Mean-Shift Contamination

Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees

A Theory for Conditional Generative Modeling on Multiple Data Sources

Provable Efficiency of Guidance in Diffusion Models for General Data Distribution

Towards Understanding Catastrophic Forgetting in Two-layer Convolutional Neural Networks

A Unified View on Learning Unnormalized Distributions via Noise-Contrastive Estimation

Grokking at the Edge of Linear Separability

Policy-Regret Minimization in Markov Games with Function Approximation

Nonlinear transformers can perform inference-time feature learning

Constrained Online Convex Optimization with Polyak Feasibility Steps

Connecting Thompson Sampling and UCB: Towards More Efficient Trade-offs Between Privacy and Regret

Prediction-Aware Learning in Multi-Agent Systems

Regret-Free Reinforcement Learning for Temporal Logic Specifications

Dueling Convex Optimization with General Preferences

Geometric Resampling in Nearly Linear Time for Follow-the-Perturbed-Leader with Best-of-Both-Worlds Guarantee in Bandit Problems

Online Episodic Convex Reinforcement Learning

On Mitigating Affinity Bias through Bandits with Evolving Biased Feedback

High-Dimensional Prediction for Sequential Decision Making

Improved Regret Analysis in Gaussian Process Bandits: Optimality for Noiseless Reward, RKHS norm, and Non-Stationary Variance

Generalization of noisy SGD in unbounded non-convex settings

A Parameter-Free and Near-Optimal Zeroth-Order Algorithm for Stochastic Convex Optimization

(ends 7:00 PM)

Poster Session 4 East [4:30-7:00]

Posters 4:30-7:00

FairICP: Encouraging Equalized Odds via Inverse Conditional Permutation

Optimal Fair Learning Robust to Adversarial Distribution Shift

Disparate Conditional Prediction in Multiclass Classifiers

KGMark: A Diffusion Watermark for Knowledge Graphs

Fairness on Principal Stratum: A New Perspective on Counterfactual Fairness

Lightweight Protocols for Distributed Private Quantile Estimation

Private Federated Learning using Preference-Optimized Synthetic Data

Privacy-Preserving Federated Convex Optimization: Balancing Partial-Participation and Efficiency via Noise Cancellation

EncryptedLLM: Privacy-Preserving Large Language Model Inference via GPU-Accelerated Fully Homomorphic Encryption

Auditing $f$-differential privacy in one run

You Get What You Give: Reciprocally Fair Federated Learning

FACTER: Fairness-Aware Conformal Thresholding and Prompt Engineering for Enabling Fair LLM-Based Recommender Systems

Fairness Overfitting in Machine Learning: An Information-Theoretic Perspective

Clone-Robust AI Alignment

Understanding Fixed Predictions via Confined Regions

Graph Inverse Style Transfer for Counterfactual Explainability

TOPLOC: A Locality Sensitive Hashing Scheme for Trustless Verifiable Inference

Evaluating Neuron Explanations: A Unified Framework with Sanity Checks

Near-Optimal Decision Trees in a SPLIT Second

X-Hacking: The Threat of Misguided AutoML

Sampling Binary Data by Denoising through Score Functions

Conditioning Diffusions Using Malliavin Calculus

Learn to Vaccinate: Combining Structure Learning and Effective Vaccination for Epidemic and Outbreak Control

Prediction-Powered E-Values

On the Vulnerability of Applying Retrieval-Augmented Generation within Knowledge-Intensive Application Domains

LIMEFLDL: A Local Interpretable Model-Agnostic Explanations Approach for Label Distribution Learning

What makes an Ensemble (Un) Interpretable?

Adversarial Cooperative Rationalization: The Risk of Spurious Correlations in Even Clean Datasets

Learning to Route LLMs with Confidence Tokens

Activation Space Interventions Can Be Transferred Between Large Language Models

Differential Privacy Guarantees of Markov Chain Monte Carlo Algorithms

New Bounds for Sparse Variational Gaussian Processes

Return of the Latent Space COWBOYS: Re-thinking the use of VAEs for Bayesian Optimisation of Structured Spaces

Determinant Estimation under Memory Constraints and Neural Scaling Laws

Outsourced Diffusion Sampling: Efficient Posterior Inference in Latent Spaces of Generative Models

Conformal Prediction as Bayesian Quadrature

Sample Complexity of Branch-length Estimation by Maximum Likelihood

A Mixture-Based Framework for Guiding Diffusion Models

LEAPS: A discrete neural sampler via locally equivariant networks

TRACE Back from the Future: A Probabilistic Reasoning Approach to Controllable Language Generation

Learning Soft Sparse Shapes for Efficient Time-Series Classification

Weakly Supervised Anomaly Detection via Dual-Tailed Kernel

Weakly-Supervised Contrastive Learning for Imprecise Class Labels

A Theoretical Framework For Overfitting In Energy-based Modeling

Partition First, Embed Later: Laplacian-Based Feature Partitioning for Refined Embedding and Visualization of High-Dimensional Data

Lightspeed Geometric Dataset Distance via Sliced Optimal Transport

Generalized Venn and Venn-Abers Calibration with Applications in Conformal Prediction

Online Detection of LLM-Generated Texts via Sequential Hypothesis Testing by Betting

Learning Survival Distributions with the Asymmetric Laplace Distribution

PTTA: Purifying Malicious Samples for Test-Time Model Adaptation

Whoever Started the interference Should End It: Guiding Data-Free Model Merging via Task Vectors

Generalized additive models via direct optimization of regularized decision stump forests

Approximately Correct Label Distribution Learning

Compressing tree ensembles through Level-wise Optimization and Pruning

Learning Imbalanced Data with Beneficial Label Noise

Predictive Performance of Deep Quantum Data Re-uploading Models

Right Time to Learn: Promoting Generalization via Bio-inspired Spacing Effect in Knowledge Distillation

Bipartite Ranking From Multiple Labels: On Loss Versus Label Aggregation

Concentration Distribution Learning from Label Distributions

Efficient LiDAR Reflectance Compression via Scanning Serialization

Improved Coresets for Vertical Federated Learning: Regularized Linear and Logistic Regressions

SKIM: Any-bit Quantization Pushing The Limits of Post-Training Quantization

Geometric Contact Flows: Contactomorphisms for Dynamics and Control

KoNODE: Koopman-Driven Neural Ordinary Differential Equations with Evolving Parameters for Time Series Analysis

Spectral-Aware Reservoir Computing for Fast and Accurate Time Series Classification

Conformal Anomaly Detection in Event Sequences

Provable Length Generalization in Sequence Prediction via Spectral Filtering

Residual TPP: A Unified Lightweight Approach for Event Stream Data Analysis

VisionTS: Visual Masked Autoencoders Are Free-Lunch Zero-Shot Time Series Forecasters

Low-distortion and GPU-compatible Tree Embeddings in Hyperbolic Space

Controlling Neural Collapse Enhances Out-of-Distribution Detection and Transfer Learning

The Complexity of Learning Sparse Superposed Features with Feedback

Contrastive Learning with Simplicial Convolutional Networks for Short-Text Classification

The Generalized Skew Spectrum of Graphs

Beyond Matryoshka: Revisiting Sparse Coding for Adaptive Representation

Near Optimal Best Arm Identification for Clustered Bandits

Wait-Less Offline Tuning and Re-solving for Online Decision Making

Efficient Core-set Selection for Deep Learning Through Squared Loss Minimization

Sample Efficient Demonstration Selection for In-Context Learning

Robust Consensus Anchor Learning for Efficient Multi-view Subspace Clustering

From Spectrum-free towards Baseline-view-free: Double-track Proximity Driven Multi-view Clustering

COKE: Core Kernel for More Efficient Approximation of Kernel Weights in Multiple Kernel Clustering

TypyBench: Evaluating LLM Type Inference for Untyped Python Repositories

Regression for the Mean: Auto-Evaluation and Inference with Few Labels through Post-hoc Regression

On Path to Multimodal Generalist: General-Level and General-Bench

Is Complex Query Answering Really Complex?

SyncMind: Measuring Agent Out-of-Sync Recovery in Collaborative Software Engineering

To Each Metric Its Decoding: Post-Hoc Optimal Decision Rules of Probabilistic Hierarchical Classifiers

Pfeife: Automatic Pipeline Parallelism for PyTorch

K$^2$IE: Kernel Method-based Kernel Intensity Estimators for Inhomogeneous Poisson Processes

Kernel Quantile Embeddings and Associated Probability Metrics

Learning Safe Strategies for Value Maximizing Buyers in Uniform Price Auctions

Clustering via Self-Supervised Diffusion

Generalization Performance of Ensemble Clustering: From Theory to Algorithm

Almost Optimal Fully Dynamic $k$-Center Clustering with Recourse

Super Deep Contrastive Information Bottleneck for Multi-modal Clustering

Fast Incomplete Multi-view Clustering by Flexible Anchor Learning

Graph Minimum Factor Distance and Its Application to Large-Scale Graph Data Clustering

Exogenous Isomorphism for Counterfactual Identifiability

Strategic A/B testing via Maximum Probability-driven Two-armed Bandit

Local Identifying Causal Relations in the Presence of Latent Variables

Distributionally Robust Policy Learning under Concept Drifts

Counterfactual Contrastive Learning with Normalizing Flows for Robust Treatment Effect Estimation

Reducing Confounding Bias without Data Splitting for Causal Inference via Optimal Transport

Differentiable Structure Learning with Ancestral Constraints

Compelling ReLU Networks to Exhibit Exponentially Many Linear Regions at Initialization and During Training

Come Together, But Not Right Now: A Progressive Strategy to Boost Low-Rank Adaptation

Feature Shift Localization Network

One-Shot Heterogeneous Federated Learning with Local Model-Guided Diffusion Models

Learning with Selectively Labeled Data from Multiple Decision-makers

Federated Learning for Feature Generalization with Convex Constraints

Disentangling Invariant Subgraph via Variance Contrastive Estimation under Distribution Shifts

Info-Coevolution: An Efficient Framework for Data Model Coevolution

Evolving Minds: Logic-Informed Inference from Temporal Action Patterns

Projection Pursuit Density Ratio Estimation

Adaptive Estimation and Learning under Temporal Distribution Shift

Provable Maximum Entropy Manifold Exploration via Diffusion Models

A Online Statistical Framework for Out-of-Distribution Detection

Advancing Constrained Monotonic Neural Networks: Achieving Universal Approximation Beyond Bounded Activations

Deep Linear Network Training Dynamics from Random Initialization: Data, Width, Depth, and Hyperparameter Transfer

Rethinking Benign Overfitting in Two-Layer Neural Networks

Constrained Belief Updates Explain Geometric Structures in Transformer Representations

A Generalizable Physics-Enhanced State Space Model for Long-Term Dynamics Forecasting in Complex Environments

LSCD: Lomb--Scargle Conditioned Diffusion for Time series Imputation

Channel Normalization for Time Series Channel Identification

Efficient Time Series Processing for Transformers and State-Space Models through Token Merging

Quantifying Memory Utilization with Effective State-Size

In-Context Fine-Tuning for Time-Series Foundation Models

Self-Supervised Learning of Intertwined Content and Positional Features for Object Detection

Non-Stationary Predictions May Be More Informative: Exploring Pseudo-Labels with a Two-Phase Pattern of Training Dynamics

Contradiction Retrieval via Contrastive Learning with Sparsity

From Local Details to Global Context: Advancing Vision-Language Models with Attention-Based Selection

Integrating Intermediate Layer Optimization and Projected Gradient Descent for Solving Inverse Problems with Diffusion Models

Learning Input Encodings for Kernel-Optimal Implicit Neural Representations

Understanding the Emergence of Multimodal Representation Alignment

One Stone, Two Birds: Enhancing Adversarial Defense Through the Lens of Distributional Discrepancy

Average Certified Radius is a Poor Metric for Randomized Smoothing

REINFORCE Adversarial Attacks on Large Language Models: An Adaptive, Distributional, and Semantic Objective

When and How Does CLIP Enable Domain and Compositional Generalization?

Phase and Amplitude-aware Prompting for Enhancing Adversarial Robustness

Long-Short Alignment for Effective Long-Context Modeling in LLMs

Simplifying DINO via Coding Rate Regularization

On the Out-of-Distribution Generalization of Self-Supervised Learning

Functional Alignment Can Mislead: Examining Model Stitching

Think Smarter not Harder: Adaptive Reasoning with Inference Aware Optimization

Taming Knowledge Conflicts in Language Models

Reinforced Lifelong Editing for Language Models

Iterative Vectors: In-Context Gradient Steering without Backpropagation

LLM Data Selection and Utilization via Dynamic Bi-level Optimization

Compress then Serve: Serving Thousands of LoRA Adapters with Little Overhead

An Analysis for Reasoning Bias of Language Models with Small Initialization

NestQuant: nested lattice quantization for matrix products and LLMs

Revolve: Optimizing AI Systems by Tracking Response Evolution in Textual Optimization

ROPO: Robust Preference Optimization for Large Language Models

Improving Your Model Ranking on Chatbot Arena by Vote Rigging

Exploiting Presentative Feature Distributions for Parameter-Efficient Continual Learning of Large Language Models

Reliable and Efficient Amortized Model-based Evaluation

AuPair: Golden Example Pairs for Code Repair

ProofAug: Efficient Neural Theorem Proving via Fine-grained Proof Structure Analysis

UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Models

Unbiased Evaluation of Large Language Models from a Causal Perspective

Reward-Guided Prompt Evolving in Reinforcement Learning for LLMs

Teaching Transformers Causal Reasoning through Axiomatic Training

KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems

Hi-Patch: Hierarchical Patch GNN for Irregular Multivariate Time Series

Learning Distribution-wise Control in Representation Space for Language Models

Progressively Label Enhancement for Large Language Model Alignment

The Geometry of Refusal in Large Language Models: Concept Cones and Representational Independence

RAPID: Long-Context Inference with Retrieval-Augmented Speculative Decoding

LLMs on the Line: Data Determines Loss-to-Loss Scaling Laws

DPO Meets PPO: Reinforced Token Optimization for RLHF

polybasic Speculative Decoding Through a Theoretical Perspective

XAttention: Block Sparse Attention with Antidiagonal Scoring

HyperIMTS: Hypergraph Neural Network for Irregular Multivariate Time Series Forecasting

Preference Adaptive and Sequential Text-to-Image Generation

xLSTM 7B: A Recurrent LLM for Fast and Efficient Inference

Hyperband-based Bayesian Optimization for Black-box Prompt Selection

SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training

BackSlash: Rate Constrained Optimized Training of Large Language Models

Scaling Sparse Feature Circuits For Studying In-Context Learning

Moirai-MoE: Empowering Time Series Foundation Models with Sparse Mixture of Experts

Improving Rationality in the Reasoning Process of Language Models through Self-playing Game

Multi-Turn Code Generation Through Single-Step Rewards

AnyEdit: Edit Any Knowledge Encoded in Language Models

Parrot: Multilingual Visual Instruction Tuning

FunBO: Discovering Acquisition Functions for Bayesian Optimization with FunSearch

AdaptiveStep: Automatically Dividing Reasoning Step through Model Confidence

Towards Lifelong Model Editing via Simulating Ideal Editor

When Bad Data Leads to Good Models

Boosting Multi-Domain Fine-Tuning of Large Language Models through Evolving Interactions between Samples

Fast Large Language Model Collaborative Decoding via Speculation

REG: Rectified Gradient Guidance for Conditional Diffusion Models

Sparsing Law: Towards Large Language Models with Greater Activation Sparsity

MASS: Mathematical Data Selection via Skill Graphs for Pretraining Large Language Models

Hidden No More: Attacking and Defending Private Third-Party LLM Inference

Discriminative Policy Optimization for Token-Level Reward Models

Benchmarking Abstract and Reasoning Abilities Through A Theoretical Perspective

Leveraging Online Olympiad-Level Math Problems for LLMs Training and Contamination-Resistant Evaluation

SkipGPT: Each Token is One of a Kind

Binary Hypothesis Testing for Softmax Models and Leverage Score Models

Structure-Guided Large Language Models for Text-to-SQL Generation

Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling

BanditSpec: Adaptive Speculative Decoding via Bandit Algorithms

R.I.P.: Better Models by Survival of the Fittest Prompts

MMInference: Accelerating Pre-filling for Long-Context Visual Language Models via Modality-Aware Permutation Sparse Attention

Ferret: Federated Full-Parameter Tuning at Scale for Large Language Models

On the Duality between Gradient Transformations and Adapters

EduLLM: Leveraging Large Language Models and Framelet-Based Signed Hypergraph Neural Networks for Student Performance Prediction

Topology-Aware Dynamic Reweighting for Distribution Shifts on Graph

On Measuring Long-Range Interactions in Graph Neural Networks

Towards LLM Unlearning Resilient to Relearning Attacks: A Sharpness-Aware Minimization Perspective and Beyond

Generalists vs. Specialists: Evaluating LLMs on Highly-Constrained Biophysical Sequence Optimization Tasks

Audio Flamingo 2: An Audio-Language Model with Long-Audio Understanding and Expert Reasoning Abilities

Oracle-MoE: Locality-preserving Routing in the Oracle Space for Memory-constrained Large Language Model Inference

Revisiting Chain-of-Thought in Code Generation: Do Language Models Need to Learn Reasoning before Coding?

Calibrated Language Models and How to Find Them with Label Smoothing

Efficient Long Context Fine-tuning with Chunk Flow

Persistent Topological Features in Large Language Models

Weisfeiler and Leman Go Gambling: Why Expressive Lottery Tickets Win

Disentangled Graph Spectral Domain Adaptation

Graph4MM: Weaving Multimodal Learning with Structural Information

Best of Both Worlds: Advantages of Hybrid Graph Sequence Models

Graph Neural Network Generalization With Gaussian Mixture Model Based Augmentation

Do We Really Need Message Passing in Brain Network Modeling?

G-Designer: Architecting Multi-agent Communication Topologies via Graph Neural Networks

Stability and Generalization Capability of Subgraph Reasoning Models for Inductive Knowledge Graph Completion

Large Language-Geometry Model: When LLM meets Equivariance

Zero-Shot Generalization of GNNs over Distinct Attribute Domains

Haste Makes Waste: A Simple Approach for Scaling Graph Neural Networks

A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition

SketchDNN: Joint Continuous-Discrete Diffusion for CAD Sketch Generation

Score-based Pullback Riemannian Geometry: Extracting the Data Manifold Geometry using Anisotropic Flows

TimeStep Master: Asymmetrical Mixture of Timestep LoRA Experts for Versatile and Efficient Diffusion Models in Vision

The Diffusion Duality

Privacy Attacks on Image AutoRegressive Models

Graph World Model

GrokFormer: Graph Fourier Kolmogorov-Arnold Transformers

TopInG: Topologically Interpretable Graph Learning via Persistent Rationale Filtration

SPHINX: Structural Prediction using Hypergraph Inference Network

Vision Graph Prompting via Semantic Low-Rank Decomposition

N2GON: Neural Networks for Graph-of-Net with Position Awareness

How Expressive are Knowledge Graph Foundation Models?

Equivalence is All: A Unified View for Self-supervised Graph Learning

Towards a Mechanistic Explanation of Diffusion Model Generalization

Wasserstein Flow Matching: Generative Modeling Over Families of Distributions

RestoreGrad: Signal Restoration Using Conditional Denoising Diffusion Models with Jointly Learned Prior

MDDM: Practical Message-Driven Generative Image Steganography Based on Diffusion Models

Smooth Interpolation for Improved Discrete Graph Generative Models

Mechanisms of Projective Composition of Diffusion Models

RocketKV: Accelerating Long-Context LLM Inference via Two-Stage KV Cache Compression

Simultaneous Multi-Robot Motion Planning with Projected Diffusion Models

Generative Audio Language Modeling with Continuous-valued Tokens and Masked Next-Token Prediction

Feynman-Kac Correctors in Diffusion: Annealing, Guidance, and Product of Experts

TabNAT: A Continuous-Discrete Joint Generative Framework for Tabular Data

Hyper-Transforming Latent Diffusion Models

Cross-regularization: Adaptive Model Complexity through Validation Gradients

RZ-NAS: Enhancing LLM-guided Neural Architecture Search via Reflective Zero-Cost Strategy

SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference

Implicit Language Models are RNNs: Balancing Parallelization and Expressivity

$\infty$-Video: A Training-Free Approach to Long Video Understanding via Continuous-Time Memory Consolidation

ParallelComp: Parallel Long-Context Compressor for Length Extrapolation

Efficient Length-Generalizable Attention via Causal Retrieval for Long-Context Language Modeling

Rethinking the Bias of Foundation Model under Long-tailed Distribution

MM-RLHF: The Next Step Forward in Multimodal LLM Alignment

Towards Graph Foundation Models: Learning Generalities Across Graphs via Task-Trees

PEAKS: Selecting Key Training Examples Incrementally via Prediction Error Anchored by Kernel Similarity

Learngene Tells You How to Customize: Task-Aware Parameter Initialization at Flexible Scales

CSTrack: Enhancing RGB-X Tracking via Compact Spatiotemporal Features

Accelerating PDE-Constrained Optimization by the Derivative of Neural Operators

Scalable Meta-Learning via Mixed-Mode Differentiation

An Efficient Matrix Multiplication Algorithm for Accelerating Inference in Binary and Ternary Neural Networks

MERGE$^3$: Efficient Evolutionary Merging on Consumer-grade GPUs

SLiM: One-shot Quantization and Sparsity with Low-rank Approximation for LLM Weight Compression

A Closer Look at Backdoor Attacks on CLIP

SBGD: Improving Graph Diffusion Generative Model via Stochastic Block Diffusion

GLGENN: A Novel Parameter-Light Equivariant Neural Networks Architecture Based on Clifford Geometric Algebras

OW-VAP: Visual Attribute Parsing for Open World Object Detection

Taming Diffusion for Dataset Distillation with High Representativeness

GS-Bias: Global-Spatial Bias Learner for Single-Image Test-Time Adaptation of Vision-Language Models

Open-Det: An Efficient Learning Framework for Open-Ended Detection

Position: We Can’t Understand AI Using our Existing Vocabulary

Position: Rethinking Explainable Machine Learning as Applied Statistics

Position: Principles of Animal Cognition to Improve LLM Evaluations

Position: The Most Expensive Part of an LLM *should* be its Training Data

Position: Political Neutrality in AI Is Impossible — But Here Is How to Approximate It

Position: Iterative Online-Offline Joint Optimization is Needed to Manage Complex LLM Copyright Risks

Tree-Sliced Wasserstein Distance: A Geometric Perspective

Position: Human Baselines in Model Evaluations Need Rigor and Transparency (With Recommendations & Reporting Checklist)

Position: Medical Large Language Model Benchmarks Should Prioritize Construct Validity

Position: AI Competitions Provide the Gold Standard for Empirical Rigor in GenAI Evaluation

Position: AI Should Not Be An Imitation Game: Centaur Evaluations

Position: All Current Generative Fidelity and Diversity Metrics are Flawed

Position: Theory of Mind Benchmarks are Broken for Large Language Models

Position: The Future of Bayesian Prediction Is Prior-Fitted

Position: We Need Responsible, Application-Driven (RAD) AI Research

Cowpox: Towards the Immunity of VLM-based Multi-Agent Systems

PoisonedEye: Knowledge Poisoning Attack on Retrieval-Augmented Generation based Large Vision-Language Models

Omni-Angle Assault: An Invisible and Powerful Physical Adversarial Attack on Face Recognition

Diffusion-based Adversarial Purification from the Perspective of the Frequency Domain

BiMark: Unbiased Multilayer Watermarking for Large Language Models

MELON: Provable Defense Against Indirect Prompt Injection Attacks in AI Agents

OR-Bench: An Over-Refusal Benchmark for Large Language Models

Improving LLM Safety Alignment with Dual-Objective Optimization

Safety Reasoning with Guidelines

Optimizing Adaptive Attacks against Watermarks for Language Models

Antidote: Post-fine-tuning Safety Alignment for Large Language Models against Harmful Fine-tuning Attack

Hardware and Software Platform Inference

BaxBench: Can LLMs Generate Correct and Secure Backends?

DIS-CO: Discovering Copyrighted Content in VLMs Training Data

Quantifying Prediction Consistency Under Fine-tuning Multiplicity in Tabular LLMs

"Who experiences large model decay and why?" A Hierarchical Framework for Diagnosing Heterogeneous Performance Drift

SCISSOR: Mitigating Semantic Bias through Cluster-Aware Siamese Networks for Robust Classification

Federated Disentangled Tuning with Textual Prior Decoupling and Visual Dynamic Adaptation

Going Deeper into Locally Differentially Private Graph Neural Networks

Scaling Test-Time Compute Without Verification or RL is Suboptimal

(ends 7:00 PM)

7 p.m.

THU 17 JUL

7:30 a.m.

Registration West

(ends 6:00 PM)

Registration East

(ends 12:00 PM)

ICML Lounge Area:

ICML Lounge Area

(ends 5:00 PM)

8:30 a.m.

Invited Talk:

What to optimize for – from robot arms to frontier AI - Anca Dragan

Anca Dragan

(ends 9:30 AM)

9:30 a.m.

Break:

Coffee Break

(ends 10:00 AM)

Exhibits

(ends 1:00 PM)

10 a.m.

Oral 5A Safety and Security [10:00-11:00]

Orals 10:00-11:00

[10:00] STAIR: Improving Safety Alignment with Introspective Reasoning

[10:15] AutoAdvExBench: Benchmarking Autonomous Exploitation of Adversarial Example Defenses

[10:30] Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards

[10:45] Model Immunization from a Condition Number Perspective

(ends 11:00 AM)

Oral 5B Deep Learning Algorithms [10:00-11:00]

Orals 10:00-11:00

[10:00] DistiLLM-2: A Contrastive Approach Boosts the Distillation of LLMs

[10:15] ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge Distillation via $\alpha$-$\beta$-Divergence

[10:30] Navigating Semantic Drift in Task-Agnostic Class-Incremental Learning

[10:45] From Weight-Based to State-Based Fine-Tuning: Further Memory Reduction on LoRA with Parallel Control

(ends 11:00 AM)

Oral 5C Probablistic Models [10:00-11:00]

Orals 10:00-11:00

[10:00] Rényi Neural Processes

[10:15] A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization

[10:30] Score Matching with Missing Data

[10:45] Beyond Self-Repellent Kernels: History-Driven Target Towards Efficient Nonlinear MCMC on General Graphs

(ends 11:00 AM)

Oral 5D Applications in Math and Physics [10:00-11:00]

Orals 10:00-11:00

[10:00] The dark side of the forces: assessing non-conservative force models for atomistic machine learning

[10:15] LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models

[10:30] Neural Discovery in Mathematics: Do Machines Dream of Colored Planes?

[10:45] Machine Learning meets Algebraic Combinatorics: A Suite of Datasets Capturing Research-level Conjecturing Ability in Pure Mathematics

(ends 11:00 AM)

Oral 5E Learning Theory [10:00-11:00]

Orals 10:00-11:00

[10:00] Statistical Query Hardness of Multiclass Linear Classification with Random Classification Noise

[10:15] All-Purpose Mean Estimation over R: Optimal Sub-Gaussianity with Outlier Robustness and Low Moments Performance

[10:30] A Generalization Result for Convergence in Learning-to-Optimize

[10:45] Theoretical Limitations of Ensembles in the Age of Overparameterization

(ends 11:00 AM)

11 a.m.

Poster Session 5 West [11:00-1:30]

Posters 11:00-1:30

Disentangling and Integrating Relational and Sensory Information in Transformer Architectures

Predicting mutational effects on protein binding from folding energy

SpikF: Spiking Fourier Network for Efficient Long-term Prediction

Local Pan-privacy for Federated Analytics

The Lock-in Hypothesis: Stagnation by Algorithm

EgoPrivacy: What Your First-Person Camera Says About You?

EnIGMA: Interactive Tools Substantially Assist LM Agents in Finding Security Vulnerabilities

Improving Generalization with Flat Hilbert Bayesian Inference

Ranked from Within: Ranking Large Multimodal Models Without Labels

Eliciting Language Model Behaviors with Investigator Agents

Monte Carlo Tree Search for Comprehensive Exploration in LLM-Based Automatic Heuristic Design

Exponential Family Variational Flow Matching for Tabular Data Generation

Quantum Speedup for Hypergraph Sparsification

Phase transitions for the existence of unregularized M-estimators in single index models

Learning With Multi-Group Guarantees For Clusterable Subpopulations

Minimax Optimal Regret Bound for Reinforcement Learning with Trajectory Feedback

General agents need world models

De-coupled NeuroGF for Shortest Path Distance Approximations on Large Terrain Graphs

Reinforcement Learning with Random Time Horizons

RepoAudit: An Autonomous LLM-Agent for Repository-Level Code Auditing

OV-MER: Towards Open-Vocabulary Multimodal Emotion Recognition

Adaptive Flow Matching for Resolving Small-Scale Physics

Zebra: In-Context Generative Pretraining for Solving Parametric PDEs

Compositional Flows for 3D Molecule and Synthesis Pathway Co-design

Geometric and Physical Constraints Synergistically Enhance Neural PDE Surrogates

Flexibility-conditioned protein structure design with flow matching

All-atom Diffusion Transformers: Unified generative modelling of molecules and materials

Closed-form Solutions: A New Perspective on Solving Differential Equations

Reinforcement Learning for Quantum Control under Physical Constraints

GenMol: A Drug Discovery Generalist with Discrete Diffusion

The dark side of the forces: assessing non-conservative force models for atomistic machine learning

LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models

Fixing the Double Penalty in Data-Driven Weather Forecasting Through a Modified Spherical Harmonic Loss Function

Riemann Tensor Neural Networks: Learning Conservative Systems with Physics-Constrained Networks

DragSolver: A Multi-Scale Transformer for Real-World Automotive Drag Coefficient Estimation

PPDiff: Diffusing in Hybrid Sequence-Structure Space for Protein-Protein Complex Design

CoastalBench: A Decade-Long High-Resolution Dataset to Emulate Complex Coastal Processes

Energy-Based Flow Matching for Generating 3D Molecular Structure

Deep Electromagnetic Structure Design Under Limited Evaluation Budgets

Learning Condensed Graph via Differentiable Atom Mapping for Reaction Yield Prediction

Counting atoms faster: policy-based nuclear magnetic resonance pulse sequencing for atomic abundance measurement

TextCenGen: Attention-Guided Text-Centric Background Adaptation for Text-to-Image Generation

Better to Teach than to Give: Domain Generalized Semantic Segmentation via Agent Queries with Diffusion Model Guidance

ExtPose: Robust and Coherent Pose Estimation by Extending ViTs

Robust Secure Swap: Responsible Face Swap With Persons of Interest Redaction and Provenance Traceability

Predicting High-precision Depth on Low-Precision Devices Using 2D Hilbert Curves

LLaVA-ReID: Selective Multi-image Questioner for Interactive Person Re-Identification

DragLoRA: Online Optimization of LoRA Adapters for Drag-based Image Editing in Diffusion Model

Improving Compositional Generation with Diffusion Models Using Lift Scores

PF3plat: Pose-Free Feed-Forward 3D Gaussian Splatting for Novel View Synthesis

The Four Color Theorem for Cell Instance Segmentation

Instruct2See: Learning to Remove Any Obstructions Across Distributions

MiraGe: Editable 2D Images using Gaussian Splatting

SSHR: More Secure Generative Steganography with High-Quality Revealed Secret Images

Not All Tokens Matter All The Time: Dynamic Token Aggregation Towards Efficient Detection Transformers

DCBM: Data-Efficient Visual Concept Bottleneck Models

LotteryCodec: Searching the Implicit Representation in a Random Network for Low-Complexity Image Compression

Flex3D: Feed-Forward 3D Generation with Flexible Reconstruction Model and Input View Curation

Origin Identification for Text-Guided Image-to-Image Diffusion Models

Mutual Learning for SAM Adaptation: A Dual Collaborative Network Framework for Source-Free Domain Transfer

Rethinking Score Distilling Sampling for 3D Editing and Generation

Diffusion on Language Model Encodings for Protein Sequence Generation

FlashTP: Fused, Sparsity-Aware Tensor Product for Machine Learning Interatomic Potentials

Unifying Specialized Visual Encoders for Video Language Models

Symmetry-Robust 3D Orientation Estimation

FlowDrag: 3D-aware Drag-based Image Editing with Mesh-guided Deformation Vector Flow Fields

RUN: Reversible Unfolding Network for Concealed Object Segmentation

The Limits of Predicting Agents from Behaviour

S2-Track: A Simple yet Strong Approach for End-to-End 3D Multi-Object Tracking

Adaptive Sensitivity Analysis for Robust Augmentation against Natural Corruptions in Image Segmentation

GoIRL: Graph-Oriented Inverse Reinforcement Learning for Multimodal Trajectory Prediction

Ultra Lowrate Image Compression with Semantic Residual Coding and Compression-aware Diffusion

Divide and Conquer: Exploring Language-centric Tree Reasoning for Video Question-Answering

NeuralCohort: Cohort-aware Neural Representation Learning for Healthcare Analytics

scSSL-Bench: Benchmarking Self-Supervised Learning for Single-Cell Data

NTPP: Generative Speech Language Modeling for Dual-Channel Spoken Dialogue via Next-Token-Pair Prediction

Flexible and Efficient Grammar-Constrained Decoding

Otter: Generating Tests from Issues to Validate SWE Patches

Are Large Language Models Ready for Multi-Turn Tabular Data Analysis?

Aligning Spoken Dialogue Models from User Interactions

SING: Spatial Context in Large Language Model for Next-Gen Wearables

OmniAudio: Generating Spatial Audio from 360-Degree Video

MindCustomer: Multi-Context Image Generation Blended with Brain Signal

Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worse

Dendritic Localized Learning: Toward Biologically Plausible Algorithm

Galileo: Learning Global & Local Features of Many Remote Sensing Modalities

Neural Discovery in Mathematics: Do Machines Dream of Colored Planes?

AlphaQCM: Alpha Discovery in Finance with Distributional Reinforcement Learning

Unbiased Recommender Learning from Implicit Feedback via Weakly Supervised Learning

AdaPTS: Adapting Univariate Foundation Models to Probabilistic Multivariate Time Series Forecasting

Cross-City Latent Space Alignment for Consistency Region Embedding

Counterfactual Voting Adjustment for Quality Assessment and Fairer Voting in Online Platforms with Helpfulness Evaluation

DynaMind: Reasoning over Abstract Video Dynamics for Embodied Decision-Making

ELEMENTAL: Interactive Learning from Demonstrations and Vision-Language Models for Reward Design in Robotics

Pre-training Auto-regressive Robotic Models with 4D Representations

Dynamical phases of short-term memory mechanisms in RNNs

EEG-Language Pretraining for Highly Label-Efficient Clinical Phenotyping

Inverse Reinforcement Learning with Switching Rewards and History Dependency for Characterizing Animal Behaviors

ReverB-SNN: Reversing Bit of the Weight and Activation for Spiking Neural Networks

A Multi-Region Brain Model to Elucidate the Role of Hippocampus in Spatially Embedded Decision-Making

The Brain's Bitter Lesson: Scaling Speech Decoding With Self-Supervised Learning

Flow Matching for Few-Trial Neural Adaptation with Stable Latent Dynamics

Identifying Neural Dynamics Using Interventional State Space Models

Human-Aligned Image Models Improve Visual Decoding from the Brain

Feature Learning beyond the Lazy-Rich Dichotomy: Insights from Representational Geometry

Not all solutions are created equal: An analytical dissociation of functional and representational similarity in deep linear neural networks

Synthesizing Images on Perceptual Boundaries of ANNs for Uncovering and Manipulating Human Perceptual Variability

MapEval: A Map-Based Evaluation of Geo-Spatial Reasoning in Foundation Models

Bridging Layout and RTL: Knowledge Distillation based Timing Prediction

Bridging Protein Sequences and Microscopy Images with Unified Diffusion Models

Machine Learning meets Algebraic Combinatorics: A Suite of Datasets Capturing Research-level Conjecturing Ability in Pure Mathematics

Feedforward Few-shot Species Range Estimation

In-Context Adaptation to Concept Drift for Learned Database Operations

DEFAME: Dynamic Evidence-based FAct-checking with Multimodal Experts

SPD: Sync-Point Drop for Efficient Tensor Parallelism of Large Language Models

Loss Functions and Operators Generated by f-Divergences

Enhancing Performance of Explainable AI Models with Constrained Concept Refinement

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

FedECADO: A Dynamical System Model of Federated Learning

Distributed Event-Based Learning via ADMM

Joint Learning of Energy-based Models and their Partition Function

A Bregman Proximal Viewpoint on Neural Operators

Multinoulli Extension: A Lossless Yet Effective Probabilistic Framework for Subset Selection over Partition Constraints

Hybrid Quantum-Classical Multi-Agent Pathfinding

Improved Theoretically-Grounded Evolutionary Algorithms for Subset Selection with a Linear Cost Constraint

Simple Randomized Rounding for Max-Min Eigenvalue Augmentation

Aequa: Fair Model Rewards in Collaborative Learning via Slimmable Networks

BSemiFL: Semi-supervised Federated Learning via a Bayesian Approach

On the Tension between Byzantine Robustness and No-Attack Accuracy in Distributed Learning

Improving Value Estimation Critically Enhances Vanilla Policy Gradient

Can We Predict Performance of Large Models across Vision-Language Tasks?

Efficient Online Reinforcement Learning for Diffusion Policy

Vintix: Action Model via In-Context Reinforcement Learning

Policy Gradient with Tree Expansion

Behavioral Exploration: Learning to Explore via In-Context Adaptation

Craftium: Bridging Flexibility and Efficiency for Rich 3D Single- and Multi-Agent Environments

Categorical Distributional Reinforcement Learning with Kullback-Leibler Divergence: Convergence and Asymptotics

Beyond The Rainbow: High Performance Deep Reinforcement Learning on a Desktop PC

Directly Forecasting Belief for Reinforcement Learning with Delays

Online Learning in Risk Sensitive constrained MDP

Rank-One Modified Value Iteration

Improved Off-policy Reinforcement Learning in Biological Sequence Design

DEALing with Image Reconstruction: Deep Attentive Least Squares

Density Ratio Estimation-based Bayesian Optimization with Semi-Supervised Learning

Adaptive Partitioning Schemes for Optimistic Optimization

Diversity By Design: Leveraging Distribution Matching for Offline Model-Based Optimization

Meta-Black-Box-Optimization through Offline Q-function Learning

A Near-Optimal Single-Loop Stochastic Algorithm for Convex Finite-Sum Coupled Compositional Optimization

Learning Curves of Stochastic Gradient Descent in Kernel Regression

Exact risk curves of signSGD in High-Dimensions: quantifying preconditioning and noise-compression effects

Cover learning for large-scale topology representation

Task-Aware Virtual Training: Enhancing Generalization in Meta-Reinforcement Learning for Out-of-Distribution Tasks

KEA: Keeping Exploration Alive by Proactively Coordinating Exploration Strategies

PIGDreamer: Privileged Information Guided World Models for Safe Partially Observable Reinforcement Learning

Learning Utilities from Demonstrations in Markov Decision Processes

O-MAPL: Offline Multi-agent Preference Learning

Learning Mean Field Control on Sparse Graphs

Learning Imperfect Information Extensive-form Games with Last-iterate Convergence under Bandit Feedback

Sleeping Reinforcement Learning

Beyond CVaR: Leveraging Static Spectral Risk Measures for Enhanced Decision-Making in Distributional Reinforcement Learning

Unveiling Markov heads in Pretrained Language Models for Offline Reinforcement Learning

Trust-Region Twisted Policy Improvement

LARM: Large Auto-Regressive Model for Long-Horizon Embodied Intelligence

Latent Imputation before Prediction: A New Computational Paradigm for De Novo Peptide Sequencing

Algorithms and Hardness for Active Learning on Graphs

Curse of High Dimensionality Issue in Transformer for Long Context Modeling

Mirror, Mirror of the Flow: How Does Regularization Shape Implicit Bias?

Consensus Is All You Get: The Role of Attention in Transformers

From Kernels to Features: A Multi-Scale Adaptive Theory of Feature Learning

Equivariant Neural Tangent Kernels

Compact Matrix Quantum Group Equivariant Neural Networks

Mind the Gap: a Spectral Analysis of Rank Collapse and Signal Propagation in Attention Layers

Gradient Flow Provably Learns Robust Classifiers for Orthonormal GMMs

Nearly Optimal Sample Complexity for Learning with Label Proportions

A Two-Stage Learning-to-Defer Approach for Multi-Task Learning

All-Purpose Mean Estimation over R: Optimal Sub-Gaussianity with Outlier Robustness and Low Moments Performance

Generation from Noisy Examples

When do neural networks learn world models?

Meta Optimality for Demographic Parity Constrained Regression via Post-Processing

Sample-Optimal Agnostic Boosting with Unlabeled Data

On the Generalization Ability of Next-Token-Prediction Pretraining

Adversarial Robustness in Two-Stage Learning-to-Defer: Algorithms and Guarantees

Provable Benefits of Unsupervised Pre-training and Transfer Learning via Single-Index Models

Rapid Overfitting of Multi-Pass SGD in Stochastic Convex Optimization

Hypothesis Testing for Generalized Thurstone Models

Statistical Query Hardness of Multiclass Linear Classification with Random Classification Noise

System-Aware Unlearning Algorithms: Use Lesser, Forget Faster

Theoretical Limitations of Ensembles in the Age of Overparameterization

Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning

A-PSRO: A Unified Strategy Learning Method with Advantage Metric for Normal-form Games

Finite-Time Convergence Rates in Stochastic Stackelberg Games with Smooth Algorithmic Agents

Best of Both Worlds: Regret Minimization versus Minimax Play

Computing Voting Rules with Improvement Feedback

Should Decision-Makers Reveal Classifiers in Online Strategic Classification?

A General Representation-Based Approach to Multi-Source Domain Adaptation

Projection Optimization: A General Framework for Multi-Objective and Multi-Group RLHF

Fluctuations of the largest eigenvalues of transformed spiked Wigner matrices

Heavy-Tailed Linear Bandits: Huber Regression with One-Pass Update

Linear Bandits with Partially Observable Features

Learning-Augmented Algorithms for MTS with Bandit Access to Multiple Predictors

Ranking with Multiple Oracles: From Weak to Strong Stochastic Transitivity

Instance-Optimal Pure Exploration for Linear Bandits on Continuous Arms

Near Optimal Non-asymptotic Sample Complexity of 1-Identification

A Classification View on Meta Learning Bandits

High Probability Bound for Cross-Learning Contextual Bandits with Unknown Context Distributions

No-Regret is not enough! Bandits with General Constraints through Adaptive Regret Minimization

Bayesian Optimization from Human Feedback: Near-Optimal Regret Bounds

Dimension-Free Adaptive Subgradient Methods with Frequent Directions

A Generalization Result for Convergence in Learning-to-Optimize

Graph-Based Algorithms for Diverse Similarity Search

Learning Mixtures of Experts with EM: A Mirror Descent Perspective

Decision Theoretic Foundations for Conformal Prediction: Optimal Uncertainty Quantification for Risk-Averse Agents

Actor-Critics Can Achieve Optimal Sample Efficiency

Do We Need to Verify Step by Step? Rethinking Process Supervision from a Theoretical Perspective

Catoni Contextual Bandits are Robust to Heavy-tailed Rewards

Logarithmic Regret for Online KL-Regularized Reinforcement Learning

A Computationally Efficient Algorithm for Infinite-Horizon Average-Reward Linear MDPs

(ends 1:30 PM)

Poster Session 5 East [11:00-1:30]

Posters 11:00-1:30

Safe Delta: Consistently Preserving Safety when Fine-Tuning LLMs on Diverse Datasets

Direct Density Ratio Optimization: A Statistically Consistent Approach to Aligning Large Language Models

Latent Preference Coding: Aligning Large Language Models via Discrete Latent Codes

Two Tickets are Better than One: Fair and Accurate Hiring Under Strategic LLM Manipulations

Accelerating Spectral Clustering under Fairness Constraints

XAttnMark: Learning Robust Audio Watermarking with Cross-Attention

Optimizing Noise Distributions for Differential Privacy

Certification for Differentially Private Prediction in Gradient-Based Training

Watch Out Your Album! On the Inadvertent Privacy Memorization in Multi-Modal Large Language Models

Efficient Source-free Unlearning via Energy-Guided Data Synthesis and Discrimination-Aware Multitask Optimization

Leveraging Sparsity for Sample-Efficient Preference Learning: A Theoretical Perspective

Mixture of Experts Made Intrinsically Interpretable

Avoiding Leakage Poisoning: Concept Interventions Under Distribution Shifts

Automatically Interpreting Millions of Features in Large Language Models

A Versatile Influence Function for Data Attribution with Non-Decomposable Loss

Selective Preference Aggregation

Jacobian Sparse Autoencoders: Sparsify Computations, Not Just Activations

Proactive Agents for Multi-Turn Text-to-Image Generation Under Uncertainty

Discovering Spoofing Attempts on Language Model Watermarks

Causal Logistic Bandits with Counterfactual Fairness Constraints

Language Models over Canonical Byte-Pair Encodings

Bayesian Inference for Correlated Human Experts and Classifiers

How to Evaluate and Mitigate IP Infringement in Visual Generative AI?

Representative Ranking for Deliberation in the Public Sphere

Trustworthy Machine Learning through Data-Specific Indistinguishability

MIB: A Mechanistic Interpretability Benchmark

Addressing Concept Mislabeling in Concept Bottleneck Models Through Preference Optimization

Are Sparse Autoencoders Useful? A Case Study in Sparse Probing

Sum-of-Parts: Self-Attributing Neural Networks with End-to-End Learning of Feature Groups

Validating Mechanistic Interpretations: An Axiomatic Approach

Memorization Sinks: Isolating Memorization during LLM Training

Variational Phylogenetic Inference with Products over Bipartitions

Continuously Updating Digital Twins using Large Language Models

Causal Discovery from Conditionally Stationary Time Series

Beyond Self-Repellent Kernels: History-Driven Target Towards Efficient Nonlinear MCMC on General Graphs

Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching

AutoStep: Locally adaptive involutive MCMC

Scaling Probabilistic Circuits via Monarch Matrices

Scalable Gaussian Processes with Latent Kronecker Structure

Stochastic Poisson Surface Reconstruction with One Solve using Geometric Gaussian Processes

Towards Robustness and Explainability of Automatic Algorithm Selection

Efficient Quantification of Multimodal Interaction at Sample Level

Editable Concept Bottleneck Models

Density Ratio Estimation with Conditional Probability Paths

Decision-aware Training of Spatiotemporal Forecasting Models to Select a Top-K Subset of Sites for Intervention

Fully Heteroscedastic Count Regression with Deep Double Poisson Networks

FedBEns: One-Shot Federated Learning based on Bayesian Ensemble

Training Flexible Models of Genetic Variant Effects from Functional Annotations using Accelerated Linear Algebra

A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization

Learning Likelihood-Free Reference Priors

Scalable Sobolev IPM for Probability Measures on a Graph

On Volume Minimization in Conformal Regression

BOOD: Boundary-based Out-Of-Distribution Data Generation

QuRe: Query-Relevant Retrieval through Hard Negative Sampling in Composed Image Retrieval

E-LDA: Toward Interpretable LDA Topic Models with Strong Guarantees in Logarithmic Parallel Time

Verification Learning: Make Unsupervised Neuro-Symbolic System Feasible

A Square Peg in a Square Hole: Meta-Expert for Long-Tailed Semi-Supervised Learning

Score Matching with Missing Data

Optimizing Temperature for Language Models with Multi-Sample Inference

DPCore: Dynamic Prompt Coreset for Continual Test-Time Adaptation

Stabilizing Sample Similarity in Representation via Mitigating Random Consistency

Robust Automatic Modulation Classification with Fuzzy Regularization

Generalized Random Forests Using Fixed-Point Trees

Surrogate Prompt Learning: Towards Efficient and Diverse Prompt Learning for Vision-Language Models

Model Immunization from a Condition Number Perspective

Predicting the Susceptibility of Examples to Catastrophic Forgetting

NegMerge: Sign-Consensual Weight Merging for Machine Unlearning

Maximum Update Parametrization and Zero-Shot Hyperparameter Transfer for Fourier Neural Operators

Set Valued Predictions For Robust Domain Generalization

Generalization Bounds via Meta-Learned Model Representations: PAC-Bayes and Sample Compression Hypernetworks

Human Cognition-Inspired Hierarchical Fuzzy Learning Machine

Trusted Multi-View Classification with Expert Knowledge Constraints

Multivariate Conformal Selection

ENSUR: Equitable and Statistically Unbiased Recommendation

Focal-SAM: Focal Sharpness-Aware Minimization for Long-Tailed Classification

Tensorized Multi-View Multi-Label Classification via Laplace Tensor Rank

Pareto-Optimal Fronts for Benchmarking Symbolic Regression Algorithms

LADA: Scalable Label-Specific CLIP Adapter for Continual Learning

Shifting Time: Time-series Forecasting with Khatri-Rao Neural Operators

Approximate Forest Completion and Learning-Augmented Algorithms for Metric Minimum Spanning Trees

TeDS: Joint Learning of Diachronic and Synchronic Perspectives in Quaternion Space for Temporal Knowledge Graph Completion

TINED: GNNs-to-MLPs by Teacher Injection and Dirichlet Energy Distillation

Learning Single Index Models with Diffusion Priors

Random Feature Representation Boosting

BDC-CLIP: Brownian Distance Covariance for Adapting CLIP to Action Recognition

Deterministic Sparse Fourier Transform for Continuous Signals with Frequency Gap

ResKoopNet: Learning Koopman Representations for Complex Dynamics with Spectral Residuals

Discovering Global False Negatives On the Fly for Self-supervised Contrastive Learning

On the Importance of Embedding Norms in Self-Supervised Learning

Self-supervised Masked Graph Autoencoder via Structure-aware Curriculum

Banyan: Improved Representation Learning with Explicit Structure

A Geometric Approach to Personalized Recommendation with Set-Theoretic Constraints Using Box Embeddings

Mitigating Local Cohesion and Global Sparseness in Graph Contrastive Learning with Fuzzy Boundaries

Mixed-curvature decision trees and random forests

Test-Time Adaptation with Binary Feedback

The Harder Path: Last Iterate Convergence for Uncoupled Learning in Zero-Sum Games with Bandit Feedback

An Optimistic Algorithm for online CMDPS with Anytime Adversarial Constraints

One-Pass Feature Evolvable Learning with Theoretical Guarantees

Bayesian Weight Enhancement with Steady-State Adaptation for Test-time Adaptation in Dynamic Environments

Distributionally Robust Active Learning for Gaussian Process Regression

Learning with Exact Invariances in Polynomial Time

Joker: Joint Optimization Framework for Lightweight Kernel Machines

An Expressive and Self-Adaptive Dynamical System for Efficient Function Learning

Raising the Bar: Investigating the Values of Large Language Models via Generative Evolving Testing

Robust ML Auditing using Prior Knowledge

WATCH: Adaptive Monitoring for AI Deployments via Weighted-Conformal Martingales

Modified K-means Algorithm with Local Optimality Guarantees

Global-Local Dirichlet Processes for Clustering Grouped Data in the Presence of Group-Specific Idiosyncratic Variables

Heterogeneous Sufficient Dimension Reduction and Subspace Clustering

Efficient Federated Incomplete Multi-View Clustering

Dynamic Similarity Graph Construction with Kernel Density Estimation

Online Sparsification of Bipartite-Like Clusters in Graphs

An Effective and Secure Federated Multi-View Clustering Method with Information-Theoretic Perspective

Correlation Clustering Beyond the Pivot Algorithm

Multi-View Graph Clustering via Node-Guided Contrastive Encoding

Hierarchical Overlapping Clustering on Graphs: Cost Function, Algorithm and Scalability

KernelBench: Can LLMs Write Efficient GPU Kernels?

Guardians of Image Quality: Benchmarking Defenses Against Adversarial Attacks on Image Quality Metrics

From Crowdsourced Data to High-quality Benchmarks: Arena-Hard and Benchbuilder Pipeline

On the Adversarial Robustness of Multi-Kernel Clustering

Since Faithfulness Fails: The Performance Limits of Neural Causal Discovery

Causal Effect Identification in lvLiNGAM from Higher-Order Cumulants

Generative Intervention Models for Causal Perturbation Modeling

FairPFN: A Tabular Foundation Model for Causal Fairness

Enhancing Treatment Effect Estimation via Active Learning: A Counterfactual Covering Perspective

Causal Attribution Analysis for Continuous Outcomes

Causal-PIK: Causality-based Physical Reasoning with a Physics-Informed Kernel

Causal Abstraction Inference under Lossy Representations

Discovering Latent Causal Graphs from Spatiotemporal Data

Transformer-Based Spatial-Temporal Counterfactual Outcomes Estimation

Learning Joint Interventional Effects from Single-Variable Interventions in Additive Models

Rethinking Causal Ranking: A Balanced Perspective on Uplift Model Evaluation

Inductive Gradient Adjustment for Spectral Bias in Implicit Neural Representations

Improving Soft Unification with Knowledge Graph Embedding Methods

Generative Human Trajectory Recovery via Embedding-Space Conditional Diffusion

SAN: Hypothesizing Long-Term Synaptic Development and Neural Engram Mechanism in Scalable Model's Parameter-Efficient Fine-Tuning

ABKD: Pursuing a Proper Allocation of the Probability Mass in Knowledge Distillation via $\alpha$-$\beta$-Divergence

Discovering a Zero (Zero-Vector Class of Machine Learning)

Volume Optimality in Conformal Prediction with Structured Prediction Sets

NTK-DFL: Enhancing Decentralized Federated Learning in Heterogeneous Settings via Neural Tangent Kernel

Importance Sampling for Nonlinear Models

Feature Importance Metrics in the Presence of Missing Data

Prediction-Powered Adaptive Shrinkage Estimation

Learning Representations of Instruments for Partial Identification of Treatment Effects

Falsification of Unconfoundedness by Testing Independence of Causal Mechanisms

Rethinking Time Encoding via Learnable Transformation Functions

LAuReL: Learned Augmented Residual Layer

Solving Probabilistic Verification Problems of Neural Networks using Branch and Bound

How to Train Your Multi-Exit Model? Analyzing the Impact of Training Strategies

Positional Attention: Expressivity and Learnability of Algorithmic Computation

Stochastic Control for Fine-tuning Diffusion Models: Optimality, Regularity, and Convergence

Risk and cross validation in ridge regression with correlated samples

Counting in Small Transformers: The Delicate Interplay between Attention and Feed-Forward Layers

Softmax is not Enough (for Sharp Size Generalisation)

Gradient Descent Converges Arbitrarily Fast for Logistic Regression via Large and Adaptive Stepsizes

Global Convergence and Rich Feature Learning in $L$-Layer Infinite-Width Neural Networks under $\mu$ Parametrization

Features are fate: a theory of transfer learning in high-dimensional regression

Understanding Input Selectivity in Mamba: Impact on Approximation Power, Memorization, and Associative Recall Capacity

Pixel-level Certified Explanations via Randomized Smoothing

SPMC: Self-Purifying Federated Backdoor Defense via Margin Contribution

Explaining the role of Intrinsic Dimensionality in Adversarial Training

CAT: Contrastive Adversarial Training for Evaluating the Robustness of Protective Perturbations in Latent Diffusion Models

Collapse-Proof Non-Contrastive Self-Supervised Learning

How does Labeling Error Impact Contrastive Learning? A Perspective from Data Dimensionality Reduction

Self-Organizing Visual Prototypes for Non-Parametric Representation Learning

TimeBridge: Non-Stationarity Matters for Long-term Time Series Forecasting

SKOLR: Structured Koopman Operator Linear RNN for Time-Series Forecasting

UnHiPPO: Uncertainty-aware Initialization for State Space Models

Retrieval Augmented Time Series Forecasting

Enhancing Foundation Models for Time Series Forecasting via Wavelet-based Tokenization

SDE Matching: Scalable and Simulation-Free Training of Latent Stochastic Differential Equations

Enhancing Certified Robustness via Block Reflector Orthogonal Layers and Logit Annealing Loss

Can DBNNs Robust to Environmental Noise for Resource-constrained Scenarios?

Putnam-AXIOM: A Functional & Static Benchmark for Measuring Higher Level Mathematical Reasoning in LLMs

Diverse Prototypical Ensembles Improve Robustness to Subpopulation Shift

Adversarial Robustness via Deformable Convolution with Stochasticity

Kernel-based Unsupervised Embedding Alignment for Enhanced Visual Representation in Vision-language Models

Interpreting CLIP with Hierarchical Sparse Autoencoders

Understanding the Limits of Deep Tabular Methods with Temporal Shift

Test-Time Learning for Large Language Models

TabFlex: Scaling Tabular Learning to Millions with Linear Attention

Objective drives the consistency of representational similarity across datasets

On the Importance of Gaussianizing Representations

Looking Beyond the Top-1: Transformers Determine Top Tokens in Order

Metadata Conditioning Accelerates Language Model Pre-training

Logits are All We Need to Adapt Closed Models

Training Software Engineering Agents and Verifiers with SWE-Gym

Cache Me If You Must: Adaptive Key-Value Quantization for Large Language Models

MA-LoT: Model-Collaboration Lean-based Long Chain-of-Thought Reasoning enhances Formal Theorem Proving

Mechanistic Unlearning: Robust Knowledge Unlearning and Editing via Mechanistic Localization

Autonomy-of-Experts Models

Scaling Laws for Floating–Point Quantization Training

QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache

Joint MoE Scaling Laws: Mixture of Experts Can Be Memory Efficient

Demystifying Singular Defects in Large Language Models

NextCoder: Robust Adaptation of Code LMs to Diverse Code Edits

Overcoming Multi-step Complexity in Multimodal Theory-of-Mind Reasoning: A Scalable Bayesian Planner

An Architecture Search Framework for Inference-Time Techniques

Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems

Deliberation in Latent Space via Differentiable Cache Augmentation

Enhancing Decision-Making of Large Language Models via Actor-Critic

Joint Localization and Activation Editing for Low-Resource Fine-Tuning

VinePPO: Refining Credit Assignment in RL Training of LLMs

Spurious Correlations in High Dimensional Regression: The Roles of Regularization, Simplicity Bias and Over-Parameterization

Overcoming Vocabulary Mismatch: Vocabulary-agnostic Teacher Guided Language Modeling

EpiCoder: Encompassing Diversity and Complexity in Code Generation

Adversarial Combinatorial Semi-bandits with Graph Feedback

Targeted Low-rank Refinement: Enhancing Sparse Language Models with Precision

Dialogue Without Limits: Constant-Sized KV Caches for Extended Response in LLMs

Optimizing Test-Time Compute via Meta Reinforcement Finetuning

Reasoning-as-Logic-Units: Scaling Test-Time Reasoning in Large Language Models Through Logic Unit Alignment

Rethinking Chain-of-Thought from the Perspective of Self-Training

Locate-then-edit for Multi-hop Factual Recall under Knowledge Editing

MAGELLAN: Metacognitive predictions of learning progress guide autotelic LLM agents in large goal spaces

InfAlign: Inference-aware language model alignment

Neutral residues: revisiting adapters for model extension

Critical Tokens Matter: Token-Level Contrastive Estimation Enhances LLM’s Reasoning Capability

An Interpretable N-gram Perplexity Threat Model for Large Language Model Jailbreaks

Speculative Prefill: Turbocharging TTFT with Lightweight and Training-Free Token Importance Estimation

Collapse or Thrive: Perils and Promises of Synthetic Data in a Self-Generating World

Language Models May Verbatim Complete Text They Were Not Explicitly Trained On

Scaling Laws for Upcycling Mixture-of-Experts Language Models

CLOVER: Cross-Layer Orthogonal Vectors Pruning

Independence Tests for Language Models

GSM-$\infty$: How Do your LLMs Behave over Infinitely Increasing Reasoning Complexity and Context Length?

The Best of Both Worlds: Bridging Quality and Diversity in Data Selection with Bipartite Graph

Compositional Generalization via Forced Rendering of Disentangled Latents

An Efficient Pruner for Large Language Model with Theoretical Guarantee

EvoMesh: Adaptive Physical Simulation with Hierarchical Graph Evolutions

OmniBal: Towards Fast Instruction-Tuning for Vision-Language Models via Omniverse Computation Balance

Cost-efficient Collaboration between On-device and Cloud Language Models

DistiLLM-2: A Contrastive Approach Boosts the Distillation of LLMs

GRAM: A Generative Foundation Reward Model for Reward Generalization

Demonstration Selection for In-Context Learning via Reinforcement Learning

Predictive Data Selection: The Data That Predicts Is the Data That Teaches

AlphaVerus: Bootstrapping Formally Verified Code Generation through Self-Improving Translation and Treefinement

Does Graph Prompt Work? A Data Operation Perspective with Theoretical Analysis

Bootstrapping Self-Improvement of Language Model Programs for Zero-Shot Schema Matching

ResearchTown: Simulator of Human Research Community

Pruning for GNNs: Lower Complexity with Comparable Expressiveness

SparseLoRA: Accelerating LLM Fine-Tuning with Contextual Sparsity

TuCo: Measuring the Contribution of Fine-Tuning to Individual Responses of LLMs

Emergent Symbolic Mechanisms Support Abstract Reasoning in Large Language Models

Heads up! Large Language Models Can Perform Tasks Without Your Instruction via Selective Attention Head Masking

Representative Language Generation

LOGO --- Long cOntext aliGnment via efficient preference Optimization

Optimizing Large Language Model Training Using FP4 Quantization

Proposer-Agent-Evaluator (PAE): Autonomous Skill Discovery For Foundation Model Internet Agents

CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration

Adapting Precomputed Features for Efficient Graph Condensation

HyperNear: Unnoticeable Node Injection Attacks on Hypergraph Neural Networks

Graph Adaptive Autoregressive Moving Average Models

LLM Enhancers for GNNs: An Analysis from the Perspective of Causal Mechanism Identification

RelGNN: Composite Message Passing for Relational Deep Learning

Beyond Message Passing: Neural Graph Pattern Machine

Beyond Topological Self-Explainable GNNs: A Formal Explainability Perspective

Hierarchical Graph Tokenization for Molecule-Language Alignment

ENAHPool: The Edge-Node Attention-based Hierarchical Pooling for Graph Neural Networks

MITIGATING OVER-EXPLORATION IN LATENT SPACE OPTIMIZATION USING LES

Improving the Diffusability of Autoencoders

I Think, Therefore I Diffuse: Enabling Multimodal In-Context Reasoning in Diffusion Models

ToMA: Token Merge with Attention for Diffusion Models

Epsilon-VAE: Denoising as Visual Decoding

Tractable Transformers for Flexible Conditional Generation

Elucidating Flow Matching ODE Dynamics via Data Geometry and Denoisers

Variational Control for Guidance in Diffusion Models

Diffusion Counterfactual Generation with Semantic Abduction

MuseControlLite: Multifunctional Music Generation with Lightweight Conditioners

EQ-VAE: Equivariance Regularized Latent Space for Improved Generative Image Modeling

Learning Robust Neural Processes with Risk-Averse Stochastic Optimization

Hessian Geometry of Latent Space in Generative Models

Improving Flow Matching by Aligning Flow Divergence

Reward-Guided Iterative Refinement in Diffusion Models at Test-Time with Applications to Protein and DNA Design

MODA: MOdular Duplex Attention for Multimodal Perception, Cognition, and Emotion Understanding

Compute Optimal Inference and Provable Amortisation Gap in Sparse Autoencoders

Rényi Neural Processes

Direct Motion Models for Assessing Generated Videos

Highly Compressed Tokenizer Can Generate Without Training

DCTdiff: Intriguing Properties of Image Generative Modeling in the DCT Space

Diffusion Sampling Correction via Approximately 10 Parameters

Editable Noise Map Inversion: Encoding Target-image into Noise For High-Fidelity Image Manipulation

Scaling Video-Language Models to 10K Frames via Hierarchical Differential Distillation

GAPrompt: Geometry-Aware Point Cloud Prompt for 3D Vision Model

LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding

Towards scientific discovery with dictionary learning: Extracting biological concepts from microscopy foundation models

Exploring Criteria of Loss Reweighting to Enhance LLM Unlearning

MUDDFormer: Breaking Residual Bottlenecks in Transformers via Multiway Dynamic Dense Connections

Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development

Distilling the Knowledge in Data Pruning

Sassha: Sharpness-aware Adaptive Second-order Optimization with Stable Hessian Approximation

Understanding the Logic of Direct Preference Alignment through Logic

Nonparametric Modern Hopfield Models

EPIC: Efficient Position-Independent Caching for Serving Large Language Models

LV-XAttn: Distributed Cross-Attention for Long Visual Inputs in Multimodal Large Language Models

MoH: Multi-Head Attention as Mixture-of-Head Attention

On the Emergence of Position Bias in Transformers

ESPFormer: Doubly-Stochastic Attention with Expected Sliced Transport Plans

Customizing the Inductive Biases of Softmax Attention using Structured Matrices

Token Coordinated Prompt Attention is Needed for Visual Prompting

How Transformers Learn Structured Data: Insights From Hierarchical Filtering

SpargeAttention: Accurate and Training-free Sparse Attention Accelerating Any Model Inference

Learning the RoPEs: Better 2D and 3D Position Encodings with STRING

Unpaired Point Cloud Completion via Unbalanced Optimal Transport

Learning from Loss Landscape: Generalizable Mixed-Precision Quantization via Adaptive Sharpness-Aware Gradient Aligning

Physics-Informed DeepONets for drift-diffusion on metric graphs: simulation and parameter identification

From Weight-Based to State-Based Fine-Tuning: Further Memory Reduction on LoRA with Parallel Control

Leveraging Per-Instance Privacy for Machine Unlearning

Navigating Semantic Drift in Task-Agnostic Class-Incremental Learning

BAnG: Bidirectional Anchored Generation for Conditional RNA Design

Stay Hungry, Keep Learning: Sustainable Plasticity for Deep Reinforcement Learning

How to Synthesize Text Data without Model Collapse?

Explicit Discovery of Nonlinear Symmetries from Dynamic Data

STAR: Learning Diverse Robot Skill Abstractions through Rotation-Augmented Vector Quantization

Core Context Aware Transformers for Long Context Language Modeling

Noise-Guided Predicate Representation Extraction and Diffusion-Enhanced Discretization for Scene Graph Generation

Position: Deep Learning is Not So Mysterious or Different

Position: The Artificial Intelligence and Machine Learning Community Should Adopt a More Transparent and Regulated Peer Review Process

Position: Build Agent Advocates, Not Platform Agents

Position: Humanity Faces Existential Risk from Gradual Disempowerment

Position: The Right to AI

Position: Strong Consumer Protection is an Inalienable Defense for AI Safety in the United States

Position: Stop treating `AGI' as the north-star goal of AI research

Adaptive Median Smoothing: Adversarial Defense for Unlearned Text-to-Image Diffusion Models at Inference Time

AutoAdvExBench: Benchmarking Autonomous Exploitation of Adversarial Example Defenses

Cannot See the Forest for the Trees: Invoking Heuristics and Biases to Elicit Irrational Choices of LLMs

Position: The AI Conference Peer Review Crisis Demands Author Feedback and Reviewer Rewards

Position: Graph Learning Will Lose Relevance Due To Poor Benchmarks

Position: Explainable AI Cannot Advance Without Better User Studies

Position: Truly Self-Improving Agents Require Intrinsic Metacognitive Learning

The Ripple Effect: On Unforeseen Complications of Backdoor Attacks

Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards

ELITE: Enhanced Language-Image Toxicity Evaluation for Safety

Improving Out-of-Distribution Detection with Markov Logic Networks

CASE-Bench: Context-Aware SafEty Benchmark for Large Language Models

Towards Black-Box Membership Inference Attack for Diffusion Models

STAIR: Improving Safety Alignment with Introspective Reasoning

Underestimated Privacy Risks for Minority Populations in Large Language Model Unlearning

Breaking the $n^{1.5}$ Additive Error Barrier for Private and Efficient Graph Sparsification via Private Expander Decomposition

Test-time Adapted Reinforcement Learning with Action Entropy Regularization

Breaking the Quadratic Barrier: Robust Cardinality Sketches for Adaptive Queries

Information Bottleneck-guided MLPs for Robust Spatial-temporal Forecasting

Scaling Trends in Language Model Robustness

Automated Red Teaming with GOAT: the Generative Offensive Agent Tester

Differentially Private Space-Efficient Algorithms for Counting Distinct Elements in the Turnstile Model

Improved Sample Complexity for Private Nonsmooth Nonconvex Optimization

Differentially Private Boxplots

Rethinking the Temperature for Federated Heterogeneous Distillation

Gradient Inversion of Multimodal Models

Faster Rates for Private Adversarial Bandits

On the Private Estimation of Smooth Transport Maps

(ends 1:30 PM)

Mentorship:

Science Communication 101: How to write an elevator pitch for your research

(ends 12:00 PM)

1 p.m.

Break:

Hosted Grab and Go Lunch

(ends 2:00 PM)

Town Hall:

Town Hall

(ends 2:00 PM)

2 p.m.

Invited Talk:

Closing the Loop: Machine Learning for Optimization and Discovery

Andreas Krause

(ends 3:00 PM)

3:30 p.m.

Oral 6A Applications in Agents and Coding [3:30-4:30]

Orals 3:30-4:30

[3:30] EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents

[3:45] SWE-Lancer: Can Frontier LLMs Earn $1 Million from Real-World Freelance Software Engineering?

[4:00] CodeIO: Condensing Reasoning Patterns via Code Input-Output Prediction

[4:15] ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Tasks

(ends 4:30 PM)

Oral 6B Deep Learning Architectures [3:30-4:30]

Orals 3:30-4:30

[3:30] Retrieval-Augmented Perception: High-resolution Image Perception Meets Visual RAG

[3:45] AutoGFM: Automated Graph Foundation Model with Adaptive Architecture Customization

[4:00] Normalizing Flows are Capable Generative Models

[4:15] In-Context Denoising with One-Layer Transformers: Connections between Attention and Associative Memory Retrieval

(ends 4:30 PM)

Oral 6C Learning Dynamics 2 [3:30-4:30]

Orals 3:30-4:30

[3:30] Learning dynamics in linear recurrent neural networks

[3:45] LoRA Training Provably Converges to a Low-Rank Global Minimum Or It Fails Loudly (But it Probably Won't Fail)

[4:00] LoRA-One: One-Step Full Gradient Could Suffice for Fine-Tuning Large Language Models, Provably and Efficiently

[4:15] Implicit Regularization for Tubal Tensor Factorizations via Gradient Descent

(ends 4:30 PM)

Oral 6D Evaluation [3:30-4:30]

Orals 3:30-4:30

[3:30] On Path to Multimodal Generalist: General-Level and General-Bench

[3:45] What Limits Virtual Agent Application? OmniBench: A Scalable Multi-Dimensional Benchmark for Essential Virtual Agent Capabilities

[4:00] How Do Large Language Monkeys Get Their Power (Laws)?

[4:15] Suitability Filter: A Statistical Framework for Classifier Evaluation in Real-World Deployment Settings

(ends 4:30 PM)

Oral 6E Social and Economic Perspectives [3:30-4:30]

Orals 3:30-4:30

[3:30] The Value of Prediction in Identifying the Worst-Off

[3:45] Generative Social Choice: The Next Generation

[4:00] Statistical Collusion by Collectives on Learning Platforms

[4:15] Prices, Bids, Values: One ML-Powered Combinatorial Auction to Rule Them All

(ends 4:30 PM)

4:30 p.m.

Break:

Coffee Break

(ends 5:30 PM)

Poster Session 6 West [4:30-7:00]

Posters 4:30-7:00

From Feature Interaction to Feature Generation: A Generative Paradigm of CTR Prediction Models

RepLoRA: Reparameterizing Low-rank Adaptation via the Perspective of Mixture of Experts

NEAR: Neural Electromagnetic Array Response

Sort Before You Prune: Improved Worst-Case Guarantees of the DiskANN Family of Graphs

Designing Cyclic Peptides via Harmonic SDE with Atom-Bond Modeling

Linear $Q$-Learning Does Not Diverge in $L^2$: Convergence Rates to a Bounded Set

Incentivize without Bonus: Provably Efficient Model-based Online Multi-agent RL for Markov Games

Convergence of Policy Mirror Descent Beyond Compatible Function Approximation

Competitively Consistent Clustering

Robust Sparsification via Sensitivity

Maximum Coverage in Turnstile Streams with Applications to Fingerprinting Measures

Polynomial-Time Approximability of Constrained Reinforcement Learning

Fixed-Confidence Multiple Change Point Identification under Bandit Feedback

No Free Lunch from Random Feature Ensembles: Scaling Laws and Near-Optimality Conditions

An Error Analysis of Flow Matching for Deep Generative Modeling

Synthesizing Software Engineering Data in a Test-Driven Manner

Provably Efficient RL for Linear MDPs under Instantaneous Safety Constraints in Non-Convex Feature Spaces

ITBench: Evaluating AI Agents across Diverse Real-World IT Automation Tasks

Towards Practical Defect-Focused Automated Code Review

POQD: Performance-Oriented Query Decomposer for Multi-vector retrieval

Deep Neural Cellular Potts Models

An All-Atom Generative Model for Designing Protein Complexes

AnalogGenie-Lite: Enhancing Scalability and Precision in Circuit Topology Discovery through Lightweight Graph Modeling

Reinforced Learning Explicit Circuit Representations for Quantum State Characterization from Local Measurements

M2PDE: Compositional Generative Multiphysics and Multi-component PDE Simulation

UniMate: A Unified Model for Mechanical Metamaterial Generation, Property Prediction, and Condition Confirmation

From Uncertain to Safe: Conformal Adaptation of Diffusion Models for Safe PDE Control

Rethink the Role of Deep Learning towards Large-scale Quantum Systems

Calibrated Physics-Informed Uncertainty Quantification

Physics-Informed Weakly Supervised Learning For Interatomic Potentials

ReQFlow: Rectified Quaternion Flow for Efficient and High-Quality Protein Backbone Generation

Unisoma: A Unified Transformer-based Solver for Multi-Solid Systems

ELoRA: Low-Rank Adaptation for Equivariant GNNs

Holistic Physics Solver: Learning PDEs in a Unified Spectral-Physical Space

SCENT: Robust Spatiotemporal Learning for Continuous Scientific Data via Scalable Conditioned Neural Fields

Semi-Supervised Blind Quality Assessment with Confidence-quantifiable Pseudo-label Learning for Authentic Images

L-Diffusion: Laplace Diffusion for Efficient Pathology Image Segmentation

Improving LLM Video Understanding with 16 Frames Per Second

What Limits Virtual Agent Application? OmniBench: A Scalable Multi-Dimensional Benchmark for Essential Virtual Agent Capabilities

Enhancing Visual Localization with Cross-Domain Image Generation

Learning Attribute-Aware Hash Codes for Fine-Grained Image Retrieval via Query Optimization

One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation

Large Displacement Motion Transfer with Unsupervised Anytime Interpolation

Staged and Physics-Grounded Learning Framework with Hyperintensity Prior for Pre-Contrast MRI Synthesis

History-Guided Video Diffusion

CAD-Editor: A Locate-then-Infill Framework with Automated Training Data Synthesis for Text-Based CAD Editing

IRBridge: Solving Image Restoration Bridge with Pre-trained Generative Diffusion Models

LightningDrag: Lightning Fast and Accurate Drag-based Image Editing Emerging from Videos

InfoSAM: Fine-Tuning the Segment Anything Model from An Information-Theoretic Perspective

ConText: Driving In-context Learning for Text Removal and Segmentation

Point Cloud Dataset Distillation

FuseUNet: A Multi-Scale Feature Fusion Method for U-like Networks

IntLoRA: Integral Low-rank Adaptation of Quantized Diffusion Models

Continual Generalized Category Discovery: Learning and Forgetting from a Bayesian Perspective

Elucidating the design space of language models for image generation

An Empirical Study on Configuring In-Context Learning Demonstrations for Unleashing MLLMs' Sentimental Perception Capability

Action Dubber: Timing Audible Actions via Inflectional Flow

MaskTwins: Dual-form Complementary Masking for Domain-Adaptive Image Segmentation

Decomposition of Graphic Design with Unified Multimodal Model

Efficient Multi-modal Long Context Learning for Training-free Adaptation

Human Body Restoration with One-Step Diffusion Model and A New Benchmark

Hierarchical Masked Autoregressive Models with Low-Resolution Token Pivots

Unlocking the Capabilities of Large Vision-Language Models for Generalizable and Explainable Deepfake Detection

DocVXQA: Context-Aware Visual Explanations for Document Question Answering

Ex-VAD: Explainable Fine-grained Video Anomaly Detection Based on Visual-Language Models

ADHMR: Aligning Diffusion-based Human Mesh Recovery via Direct Preference Optimization

BiMaCoSR: Binary One-Step Diffusion Model Leveraging Flexible Matrix Compression for Real Super-Resolution

Confounder-Free Continual Learning via Recursive Feature Normalization

AffinityFlow: Guided Flows for Antibody Affinity Maturation

SToFM: a Multi-scale Foundation Model for Spatial Transcriptomics

H-Tuning: Toward Low-Cost and Efficient ECG-based Cardiovascular Disease Detection with Pre-Trained Models

Multimodal Medical Code Tokenizer

HealthGPT: A Medical Large Vision-Language Model for Unifying Comprehension and Generation via Heterogeneous Knowledge Adaptation

An Online Adaptive Sampling Algorithm for Stochastic Difference-of-convex Optimization with Time-varying Distributions

From Token to Rhythm: A Multi-Scale Approach for ECG-Language Pretraining

Global Context-aware Representation Learning for Spatially Resolved Transcriptomics

Distributed Parallel Gradient Stacking(DPGS): Solving Whole Slide Image Stacking Challenge in Multi-Instance Learning

P(all-atom) Is Unlocking New Path For Protein Design

Active Evaluation Acquisition for Efficient LLM Benchmarking

Overcoming Non-monotonicity in Transducer-based Streaming Generation

From Complex to Atomic: Enhancing Augmented Generation via Knowledge-Aware Dual Rewriting and Reasoning

ALMTokenizer: A Low-bitrate and Semantic-rich Audio Codec Tokenizer for Audio Language Modeling

KBQA-o1: Agentic Knowledge Base Question Answering with Monte Carlo Tree Search

Flow Matching for Denoised Social Recommendation

EvoControl: Multi-Frequency Bi-Level Control for High-Frequency Continuous Control

Learning Efficient Robotic Garment Manipulation with Standardization

Hi Robot: Open-Ended Instruction Following with Hierarchical Vision-Language-Action Models

Falcon: Fast Visuomotor Policies via Partial Denoising

Learning Safe Control via On-the-Fly Bandit Exploration

UP-VLA: A Unified Understanding and Prediction Model for Embodied Agent

BiAssemble: Learning Collaborative Affordance for Bimanual Geometric Assembly

Rethinking Latent Redundancy in Behavior Cloning: An Information Bottleneck Approach for Robot Manipulation

ReinboT: Amplifying Robot Visual-Language Manipulation with Reinforcement Learning

One-Step Diffusion Policy: Fast Visuomotor Policies via Diffusion Distillation

DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning

TeLoGraF: Temporal Logic Planning via Graph-encoded Flow Matching

SpikeVideoFormer: An Efficient Spike-Driven Video Transformer with Hamming Attention and $\mathcal{O}(T)$ Complexity

Neural Encoding and Decoding at Scale

Contour Integration Underlies Human-Like Vision

CogReact: A Reinforced Framework to Model Human Cognitive Reaction Modulated by Dynamic Intervention

Revisiting Noise Resilience Strategies in Gesture Recognition: Short-Term Enhancement in sEMG Analysis

OWLS: Scaling Laws for Multilingual Speech Recognition and Translation Models

Grammar-Forced Translation of Natural Language to Temporal Logic using LLMs

DMOSpeech: Direct Metric Optimization via Distilled Diffusion Model in Zero-Shot Speech Synthesis

Reducing Tool Hallucination via Reliability Alignment

FIC-TSC: Learning Time Series Classification with Fisher Information Constraint

ITFormer: Bridging Time Series and Natural Language for Multi-Modal QA with Large-Scale Multitask Dataset

CFPT: Empowering Time Series Forecasting through Cross-Frequency Interaction and Periodic-Aware Timestamp Modeling

FSTLLM: Spatio-Temporal LLM for Few Shot Time Series Forecasting

MARS: Unleashing the Power of Variance Reduction for Training Large Models

A Non-isotropic Time Series Diffusion Model with Moving Average Transitions

CVE-Bench: A Benchmark for AI Agents’ Ability to Exploit Real-World Web Application Vulnerabilities

Targeted control of fast prototyping through domain-specific interface

CodeSync: Synchronizing Large Language Models with Dynamic Code Evolution at Scale

Curriculum Learning for Biological Sequence Prediction: The Case of De Novo Peptide Sequencing

DeepLayout: Learning Neural Representations of Circuit Placement Layout

Learning Cascade Ranking as One Network

Alpha-SQL: Zero-Shot Text-to-SQL using Monte Carlo Tree Search

PatchPilot: A Cost-Efficient Software Engineering Agent with Early Attempts on Formal Verification

End-to-End Learning Framework for Solving Non-Markovian Optimal Control

Optimization over Sparse Support-Preserving Sets: Two-Step Projection with Global Optimality Guarantees

Graph-Supported Dynamic Algorithm Configuration for Multi-Objective Combinatorial Optimization

Triple-Optimistic Learning for Stochastic Contextual Bandits with General Constraints

Embedding Safety into RL: A New Take on Trust Region Methods

Revisiting Convergence: Shuffling Complexity Beyond Lipschitz Smoothness

Improved Last-Iterate Convergence of Shuffling Gradient Methods for Nonsmooth Convex Optimization

Improving Reward Model Generalization from Adversarial Process Enhanced Preferences

Reinforcement Learning with Segment Feedback

Learning to Reuse Policies in State Evolvable Environments

Guided Zeroth-Order Methods for Stochastic Non-convex Problems with Decision-Dependent Distributions

Natural Perturbations for Black-box Training of Neural Networks by Zeroth-Order Optimization

Global Optimization with a Power-Transformed Objective and Gaussian Smoothing

TSP: A Two-Sided Smoothed Primal-Dual Method for Nonconvex Bilevel Optimization

Generalized Smooth Bilevel Optimization with Nonconvex Lower-Level

Hybrid Batch Normalisation: Resolving the Dilemma of Batch Normalisation in Federated Learning

Demystifying Cost-Efficiency in LLM Serving over Heterogeneous GPUs

QoS-Efficient Serving of Multiple Mixture-of-Expert LLMs Using Partial Runtime Reconfiguration

DSP: Dynamic Sequence Parallelism for Multi-Dimensional Transformers

Unified Breakdown Analysis for Byzantine Robust Gossip

Layer-wise Quantization for Quantized Optimistic Dual Averaging

Provably Near-Optimal Federated Ensemble Distillation with Negligible Overhead

Large Language Model-driven Large Neighborhood Search for Large-Scale MILP Problems

Neural Solver Selection for Combinatorial Optimization

Don't Restart, Just Reuse: Reoptimizing MILPs with Dynamic Parameters

Preference Optimization for Combinatorial Optimization Problems

A Mixed-Curvature based Pre-training Paradigm for Multi-Task Vehicle Routing Solver

BOPO: Neural Combinatorial Optimization via Best-anchored and Objective-guided Preference Optimization

Extreme Value Policy Optimization for Safe Reinforcement Learning

Offline-to-Online Reinforcement Learning with Classifier-Free Diffusion Generation

Reward Translation via Reward Machine in Semi-Alignable MDPs

Catching Two Birds with One Stone: Reward Shaping with Dual Random Networks for Balancing Exploration and Exploitation

An Online Learning Approach to Prompt-based Selection of Generative Models and LLMs

Hyper: Hyperparameter Robust Efficient Exploration in Reinforcement Learning

Reward-free World Models for Online Imitation Learning

Quantum Algorithms for Finite-horizon Markov Decision Processes

Conservative Offline Goal-Conditioned Implicit V-Learning

Graph-Assisted Stitching for Offline Hierarchical Reinforcement Learning

C2IQL: Constraint-Conditioned Implicit Q-learning for Safe Offline Reinforcement Learning

MODULI: Unlocking Preference Generalization via Diffusion Models for Offline Multi-Objective Reinforcement Learning

Ensemble Learned Bloom Filters: Two Oracles are Better than One

Video-Enhanced Offline Reinforcement Learning: A Model-Based Approach

Score-Based Diffusion Policy Compatible with Reinforcement Learning via Optimal Transport

Learning from Suboptimal Data in Continuous Control via Auto-Regressive Soft Q-Network

FOUNDER: Grounding Foundation Models in World Models for Open-Ended Embodied Decision Making

A Forget-and-Grow Strategy for Deep Reinforcement Learning Scaling in Continuous Control

Optimizing Language Models for Inference Time Objectives using Reinforcement Learning

Behavior-agnostic Task Inference for Robust Offline In-context Reinforcement Learning

Deep Reinforcement Learning from Hierarchical Preference Design

R*: Efficient Reward Design via Reward Structure Evolution and Parameter Alignment Optimization with Large Language Models

Learning dynamics in linear recurrent neural networks

Understanding Nonlinear Implicit Bias via Region Counts in Input Space

On the Benefits of Active Data Collection in Operator Learning

Optimistic Algorithms for Adaptive Estimation of the Average Treatment Effect

Lower Bounds for Chain-of-Thought Reasoning in Hard-Attention Transformers

Conformal Tail Risk Control for Large Language Model Alignment

Improved Approximations for Hard Graph Problems using Predictions

Speeding up Policy Simulation in Supply Chain RL

Convergence Analysis of Policy Gradient Methods with Dynamic Stochasticity

Rejecting Hallucinated State Targets during Planning

Simple Policy Optimization

Ad Hoc Teamwork via Offline Goal-Based Decision Transformers

DipLLM: Fine-Tuning LLM for Strategic Decision-making in Diplomacy

Robust Multi-Agent Reinforcement Learning with Stochastic Adversary

LLM-Assisted Semantically Diverse Teammate Generation for Efficient Multi-agent Coordination

PokéChamp: an Expert-level Minimax Language Agent

Learning Strategic Language Agents in the Werewolf Game with Iterative Latent Space Policy Optimization

Concurrent Reinforcement Learning with Aggregated States via Randomized Least Squares Value Iteration

GradPS: Resolving Futile Neurons in Parameter Sharing Network for Multi-Agent Reinforcement Learning

M³HF: Multi-agent Reinforcement Learning from Multi-phase Human Feedback of Mixed Quality

Decoding Rewards in Competitive Games: Inverse Game Theory with Entropy Regularization

Robust Autonomy Emerges from Self-Play

Measuring Representational Shifts in Continual Learning: A Linear Transformation Perspective

Low-Dimension-to-High-Dimension Generalization and Its Implications for Length Generalization

Runtime Analysis of Evolutionary NAS for Multiclass Classification

On the Role of Label Noise in the Feature Learning Process

LoRA Training Provably Converges to a Low-Rank Global Minimum Or It Fails Loudly (But it Probably Won't Fail)

Understanding Overadaptation in Supervised Fine-Tuning: The Role of Ensemble Methods

Theoretical Performance Guarantees for Partial Domain Adaptation via Partial Optimal Transport

When can in-context learning generalize out of task distribution?

Transfer Learning for Nonparametric Contextual Dynamic Pricing

Prices, Bids, Values: One ML-Powered Combinatorial Auction to Rule Them All

Generative Social Choice: The Next Generation

Safely Learning Optimal Auctions: A Testable Learning Framework for Mechanism Design

Procurement Auctions via Approximately Optimal Submodular Optimization

Observation Interference in Partially Observable Assistance Games

Implicit Regularization for Tubal Tensor Factorizations via Gradient Descent

LASER: Attention with Exponential Transformation

A New Concentration Inequality for Sampling Without Replacement and Its Application for Transductive Learning

Refined generalization analysis of the Deep Ritz Method and Physics-Informed Neural Networks

Generalization in Federated Learning: A Conditional Mutual Information Framework

Exactly Tight Information-theoretic Generalization Bounds via Binary Jensen-Shannon Divergence

Models of Heavy-Tailed Mechanistic Universality

Tight and Fast Bounds for Multi-Label Learning

(ends 7:00 PM)

Poster Session 6 East [4:30-7:00]

Posters 4:30-7:00

Understanding the difficulties of posterior predictive estimation

Deep Bayesian Filter for Bayes-Faithful Data Assimilation

Splitting & Integrating: Out-of-Distribution Detection via Adversarial Gradient Attribution

Low-Rank Adapting Models for Sparse Autoencoders

B-score: Detecting biases in large language models using response history

How Do Transformers Learn Variable Binding in Symbolic Programs?

Towards flexible perception with visual memory

InfoCons: Identifying Interpretable Critical Concepts in Point Clouds via Information Theory

SAE-V: Interpreting Multimodal Models for Enhanced Alignment

CoDy: Counterfactual Explainers for Dynamic Graphs

Weight matrices compression based on PDB model in deep neural networks

Solving Linear-Gaussian Bayesian Inverse Problems with Decoupled Diffusion Sequential Monte Carlo

Learning Gaussian DAG Models without Condition Number Bounds

Parallel Simulation for Log-concave Sampling and Score-based Diffusion Models

Sampling from Binary Quadratic Distributions via Stochastic Localization

Enabling Optimal Decisions in Rehearsal Learning under CARE Condition

Trajectory Inference with Smooth Schrödinger Bridges

Neighbour-Driven Gaussian Process Variational Autoencoders for Scalable Structured Latent Modelling

Optimal Sensor Scheduling and Selection for Continuous-Discrete Kalman Filtering with Auxiliary Dynamics

Posterior Inference with Diffusion Models for High-dimensional Black-box Optimization

Generalized Category Discovery via Reciprocal Learning and Class-Wise Distribution Regularization

Unsupervised Learning for Class Distribution Mismatch

Learning Dynamics under Environmental Constraints via Measurement-Induced Bundle Structures

Online Differentially Private Conformal Prediction for Uncertainty Quantification

Large Continual Instruction Assistant

On the Diversity of Adversarial Ensemble Learning

Efficient Bisection Projection to Ensure Neural-Network Solution Feasibility for Optimization over General Set

Probabilistic Group Mask Guided Discrete Optimization for Incremental Learning

Learning from True-False Labels via Multi-modal Prompt Retrieving

Random Policy Evaluation Uncovers Policies of Generative Flow Networks

Towards the Efficient Inference by Incorporating Automated Computational Phenotypes under Covariate Shift

Socialized Coevolution: Advancing a Better World through Cross-Task Collaboration

Upweighting Easy Samples in Fine-Tuning Mitigates Forgetting

Fast and Robust: Task Sampling with Posterior and Diversity Synergies for Adaptive Decision-Makers in Randomized Environments

Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent

BSLoRA: Enhancing the Parameter Efficiency of LoRA with Intra-Layer and Inter-Layer Sharing

WeGeFT: Weight‑Generative Fine‑Tuning for Multi‑Faceted Efficient Adaptation of Large Models

Self-Bootstrapping for Versatile Test-Time Adaptation

Learnable Spatial-Temporal Positional Encoding for Link Prediction

Sparse Causal Discovery with Generative Intervention for Unsupervised Graph Domain Adaptation

Improving Multi-Class Calibration through Normalization-Aware Isotonic Techniques

TLLC: Transfer Learning-based Label Completion for Crowdsourcing

Cut out and Replay: A Simple yet Versatile Strategy for Multi-Label Online Continual Learning

Gradient Aligned Regression via Pairwise Losses

Learnware Specification via Dual Alignment

Scalable Model Merging with Progressive Layer-wise Distillation

Zero-shot Meta-learning for Tabular Prediction Tasks with Adversarially Pre-trained Transformer

When Data-Free Knowledge Distillation Meets Non-Transferable Teacher: Escaping Out-of-Distribution Trap is All You Need

Representation Surgery in Model Merging with Probabilistic Modeling

CABS: Conflict-Aware and Balanced Sparsification for Enhancing Model Merging

Learning Compact Semantic Information for Incomplete Multi-View Missing Multi-Label Classification

Towards Escaping from Class Dependency Modeling for Multi-Dimensional Classification

Preserving AUC Fairness in Learning with Noisy Protected Groups

KAN-AD: Time Series Anomaly Detection with Kolmogorov–Arnold Networks

TreeLoRA: Efficient Continual Learning via Layer-Wise LoRAs Guided by a Hierarchical Gradient-Similarity Tree

Slimming the Fat-Tail: Morphing-Flow for Adaptive Time Series Modeling

TimeDART: A Diffusion Autoregressive Transformer for Self-Supervised Time Series Representation

DTZO: Distributed Trilevel Zeroth Order Learning with Provable Non-Asymptotic Convergence

Efficient Network Automatic Relevance Determination

Balancing Model Efficiency and Performance: Adaptive Pruner for Long-tailed Data

OrcaLoca: An LLM Agent Framework for Software Issue Localization

Nested Expectations with Kernel Quadrature

An Instrumental Value for Data Production and its Application to Data Pricing

On the Dynamic Regret of Following the Regularized Leader: Optimism with History Pruning

AutoAL: Automated Active Learning with Differentiable Query Strategy Search

Improved and Oracle-Efficient Online $\ell_1$-Multicalibration

Online Clustering of Dueling Bandits

SDMG: Smoothing Your Diffusion Models for Powerful Graph Representation Learning

TraceGrad: a Framework Learning Expressive SO(3)-equivariant Non-linear Representations for Electronic-Structure Hamiltonian Prediction

FEAT-KD: Learning Concise Representations for Single and Multi-Target Regression via TabNet Knowledge Distillation

AKRMap: Adaptive Kernel Regression for Trustworthy Visualization of Cross-Modal Embeddings

How Do Large Language Monkeys Get Their Power (Laws)?

AAAR-1.0: Assessing AI’s Potential to Assist Research

Do Vision-Language Models Really Understand Visual Language?

Reflection-Bench: Evaluating Epistemic Agency in Large Language Models

Minerva: A Programmable Memory Test Benchmark for Language Models

High-Dimensional Tensor Regression With Oracle Properties

SWE-Lancer: Can Frontier LLMs Earn $1 Million from Real-World Freelance Software Engineering?

Federated Node-Level Clustering Network with Cross-Subgraph Link Mending

A Peer-review Look on Multi-modal Clustering: An Information Bottleneck Realization Method

Generalizing Causal Effects from Randomized Controlled Trials to Target Populations across Diverse Environments

Unraveling the Interplay between Carryover Effects and Reward Autocorrelations in Switchback Experiments

Counterfactual Graphical Models: Constraints and Inference

Latent Variable Causal Discovery under Selection Bias

Bivariate Causal Discovery with Proxy Variables: Integral Solving and Beyond

Federated Causal Structure Learning with Non-identical Variable Sets

LLMScan: Causal Scan for LLM Misbehavior Detection

Causal Abstraction Learning based on the Semantic Embedding Principle

Measuring Variable Importance in Heterogeneous Treatment Effects with Confidence

TANGO: Clustering with Typicality-Aware Nonlocal Mode-Seeking and Graph-Cut Optimization

Learning from Sample Stability for Deep Clustering

Unified K-Means Clustering with Label-Guided Manifold Learning

PROTOCOL: Partial Optimal Transport-enhanced Contrastive Learning for Imbalanced Multi-view Clustering

Variational Counterfactual Intervention Planning to Achieve Target Outcomes

Bayesian Active Learning for Bivariate Causal Discovery

FedClean: A General Robust Label Noise Correction for Federated Learning

OpenworldAUC: Towards Unified Evaluation and Optimization for Open-world Prompt Tuning

Understanding Model Reprogramming for CLIP via Decoupling Visual Prompts

Testing Conditional Mean Independence Using Generative Neural Networks

Autoencoder-Based Hybrid Replay for Class-Incremental Learning

Parametric Scaling Law of Tuning Bias in Conformal Prediction

Conformal Prediction with Cellwise Outliers: A Detect-then-Impute Approach

Supercharging Graph Transformers with Advective Diffusion

When Dynamic Data Selection Meets Data Augmentation: Achieving Enhanced Training Acceleration

MME-CoT: Benchmarking Chain-of-Thought in Large Multimodal Models for Reasoning Quality, Robustness, and Efficiency

Efficient Logit-based Knowledge Distillation of Deep Spiking Neural Networks for Full-Range Timestep Deployment

WAVE: Weighted Autoregressive Varying Gate for Time Series Forecasting

TimeStacker: A Novel Framework with Multilevel Observation for Capturing Nonstationary Patterns in Time Series Forecasting

TimeFilter: Patch-Specific Spatial-Temporal Graph Filtration for Time Series Forecasting

Implicit Bias of Gradient Descent for Non-Homogeneous Deep Networks

Simplicity Bias and Optimization Threshold in Two-Layer ReLU Networks

Task Generalization with Autoregressive Compositional Structure: Can Learning from $D$ Tasks Generalize to $D^T$ Tasks?

Sharp Generalization for Nonparametric Regression by Over-Parameterized Neural Networks: A Distribution-Free Analysis in Spherical Covariate

Understanding Sharpness Dynamics in NN Training with a Minimalist Example: The Effects of Dataset Difficulty, Depth, Stochasticity, and More

Tree-Sliced Wasserstein Distance with Nonlinear Projection

Widening the Network Mitigates the Impact of Data Heterogeneity on FedAvg

Sketch to Adapt: Fine-Tunable Sketches for Efficient LLM Adaptation

LaCache: Ladder-Shaped KV Caching for Efficient Long-Context Modeling of Large Language Models

Linear Mode Connectivity between Multiple Models modulo Permutation Symmetries

Linear Transformers as VAR Models: Aligning Autoregressive Attention Mechanisms with Autoregressive Forecasting

Measuring In-Context Computation Complexity via Hidden State Prediction

Time-VLM: Exploring Multimodal Vision-Language Models for Augmented Time Series Forecasting

A Closer Look at Transformers for Time Series Forecasting: Understanding Why They Work and Where They Struggle

ML$^2$-GCL: Manifold Learning Inspired Lightweight Graph Contrastive Learning

Clustering Properties of Self-Supervised Learning

M3-JEPA: Multimodal Alignment via Multi-gate MoE based on the Joint-Embedding Predictive Architecture

Structure-informed Risk Minimization for Robust Ensemble Learning

Adversarial Inputs for Linear Algebra Backends

Improving Zero-Shot Adversarial Robustness in Vision-Language Models by Closed-form Alignment of Adversarial Path Simplices

ROME is Forged in Adversity: Robust Distilled Datasets via Information Bottleneck

You Always Recognize Me (YARM): Robust Texture Synthesis Against Multi-View Corruption

Geometric Median (GM) Matching for Robust k-Subset Selection from Noisy Data

HaploVL: A Single-Transformer Baseline for Multi-Modal Understanding

MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems

Reward Modeling with Ordinal Feedback: Wisdom of the Crowd

The Berkeley Function Calling Leaderboard (BFCL): From Tool Use to Agentic Evaluation of Large Language Models

La RoSA: Enhancing LLM Efficiency via Layerwise Rotated Sparse Activation

Data Mixing Optimization for Supervised Fine-Tuning of Large Language Models

Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding

unMORE: Unsupervised Multi-Object Segmentation via Center-Boundary Reasoning

Controllable Data Generation with Hierarchical Neural Representations

ActionPiece: Contextually Tokenizing Action Sequences for Generative Recommendation

Reaction Graph: Towards Reaction-Level Modeling for Chemical Reactions with 3D Structures

TtBA: Two-third Bridge Approach for Decision-Based Adversarial Attack

DiffAdvMAP: Flexible Diffusion-Based Framework for Generating Natural Unrestricted Adversarial Examples

HyperTree Planning: Enhancing LLM Reasoning via Hierarchical Thinking

Everything Everywhere All at Once: LLMs can In-Context Learn Multiple Tasks in Superposition

Earley-Driven Dynamic Pruning for Efficient Structured Decoding

Imagine While Reasoning in Space: Multimodal Visualization-of-Thought

PENCIL: Long Thoughts with Short Memory

Mask-Enhanced Autoregressive Prediction: Pay Less Attention to Learn More

TruthFlow: Truthful LLM Generation via Representation Flow Correction

MPO: An Efficient Post-Processing Framework for Mixing Diverse Preference Alignment

GIVE: Structured Reasoning of Large Language Models with Knowledge Graph Inspired Veracity Extrapolation

Memory Layers at Scale

Model Swarms: Collaborative Search to Adapt LLM Experts via Swarm Intelligence

Test-Time Preference Optimization: On-the-Fly Alignment via Iterative Textual Feedback

GraphGPT: Generative Pre-trained Graph Eulerian Transformer

SpeCache: Speculative Key-Value Caching for Efficient Generation of LLMs

Learning to Plan & Reason for Evaluation with Thinking-LLM-as-a-Judge

Demystifying Long Chain-of-Thought Reasoning

Compositional Causal Reasoning Evaluation in Language Models

Policy Guided Tree Search for Enhanced LLM Reasoning

Do NOT Think That Much for 2+3=? On the Overthinking of Long Reasoning Models

CoMemo: LVLMs Need Image Context with Image Memory

Mixture of Hidden-Dimensions: Not All Hidden-States’ Dimensions are Needed in Transformer

Benign Samples Matter! Fine-tuning On Outlier Benign Samples Severely Breaks Safety

AlphaDPO: Adaptive Reward Margin for Direct Preference Optimization

GRU: Mitigating the Trade-off between Unlearning and Retention for LLMs

EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents

Hypo3D: Exploring Hypothetical Reasoning in 3D

Robust Multimodal Large Language Models Against Modality Conflict

DocKS-RAG: Optimizing Document-Level Relation Extraction through LLM-Enhanced Hybrid Prompt Tuning

Teaching Language Models to Critique via Reinforcement Learning

Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models

Gravity-Bench-v1: A Benchmark on Gravitational Physics Discovery for Agents

Interpreting the Repeated Token Phenomenon in Large Language Models

Non-Asymptotic Length Generalization

What Do Learning Dynamics Reveal About Generalization in LLM Mathematical Reasoning?

Overtrained Language Models Are Harder to Fine-Tune

Perception in Reflection

CROW: Eliminating Backdoors from Large Language Models via Internal Consistency Regularization

Resolving Lexical Bias in Model Editing

Programming Every Example: Lifting Pre-training Data Quality Like Experts at Scale

Retraining-free Merging of Sparse MoE via Hierarchical Clustering

C-3PO: Compact Plug-and-Play Proxy Optimization to Achieve Human-like Retrieval-Augmented Generation

CodeIO: Condensing Reasoning Patterns via Code Input-Output Prediction

QLASS: Boosting Language Agent Inference via Q-Guided Stepwise Search

CodeSteer: Symbolic-Augmented Language Models via Code/Text Guidance

Modularized Self-Reflected Video Reasoner for Multimodal LLM with Application to Video Question Answering

Product of Experts with LLMs: Boosting Performance on ARC Is a Matter of Perspective

Understanding Synthetic Context Extension via Retrieval Heads

Streamline Without Sacrifice - Squeeze out Computation Redundancy in LMM

Tuning LLM Judge Design Decisions for 1/1000 of the Cost

EVOLvE: Evaluating and Optimizing LLMs For In-Context Exploration

Controlling Large Language Model with Latent Action

Language Models as Implicit Tree Search

Automated Hypothesis Validation with Agentic Sequential Falsifications

What Limits Bidirectional Model's Generative Capabilities? A Uni-Bi-Directional Mixture-of-Expert Method For Bidirectional Fine-tuning

Rethinking External Slow-Thinking: From Snowball Errors to Probability of Correct Reasoning

Accelerating Large Language Model Reasoning via Speculative Search

GANQ: GPU-Adaptive Non-Uniform Quantization for Large Language Models

Diverging Preferences: When do Annotators Disagree and do Models Know?

On Teacher Hacking in Language Model Distillation

MoRAgent: Parameter Efficient Agent Tuning with Mixture-of-Roles

Reward-Augmented Data Enhances Direct Preference Alignment of LLMs

T1: Advancing Language Model Reasoning through Reinforcement Learning and Inference Scaling

LLMs Can Reason Faster Only If We Let Them

SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning

RULEBREAKERS: Challenging LLMs at the Crossroads between Formal Logic and Human-like Reasoning

Statistical Hypothesis Testing for Auditing Robustness in Language Models

Thickness-aware E(3)-Equivariant 3D Mesh Neural Networks

Adaptive Message Passing: A General Framework to Mitigate Oversmoothing, Oversquashing, and Underreaching

EAGLES: Towards Effective, Efficient, and Economical Federated Graph Learning via Unified Sparsification

GPEN: Global Position Encoding Network for Enhanced Subgraph Representation Learning

Open Your Eyes: Vision Enhances Message Passing Neural Networks in Link Prediction

Positional Encoding meets Persistent Homology on Graphs

PROXSPARSE: REGULARIZED LEARNING OF SEMI-STRUCTURED SPARSITY MASKS FOR PRETRAINED LLMS

STP: Self-play LLM Theorem Provers with Iterative Conjecturing and Proving

Let LLM Tell What to Prune and How Much to Prune

Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks

FedPHA: Federated Prompt Learning for Heterogeneous Client Adaptation

The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions

From Theory to Practice: Rethinking Green and Martin Kernels for Unleashing Graph Transformers

AutoGFM: Automated Graph Foundation Model with Adaptive Architecture Customization

Directed Graph Grammars for Sequence-based Learning

WILTing Trees: Interpreting the Distance Between MPNN Embeddings

Energy-Based Preference Model Offers Better Offline Alignment than the Bradley-Terry Preference Model

Expressive Power of Graph Neural Networks for (Mixed-Integer) Quadratic Programs

Delay-DSGN: A Dynamic Spiking Graph Neural Network with Delay Mechanisms for Evolving Graph

PieClam: A Universal Graph Autoencoder Based on Overlapping Inclusive and Exclusive Communities

Covered Forest: Fine-grained generalization analysis of graph neural networks

MimicMotion: High-Quality Human Motion Video Generation with Confidence-aware Pose Guidance

SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer

Normalizing Flows are Capable Generative Models

Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflow

SongGen: A Single Stage Auto-regressive Transformer for Text-to-Song Generation

Introducing 3D Representation for Dense Volume-to-Volume Translation via Score Fusion

ZipAR: Parallel Autoregressive Image Generation through Spatial Locality

MissScore: High-Order Score Estimation in the Presence of Missing Data

Training Diffusion-based Generative Models with Limited Data

AdvI2I: Adversarial Image Attack on Image-to-Image Diffusion Models

Graph Generative Pre-trained Transformer

Inverse Flow and Consistency Models

CHATS: Combining Human-Aligned Optimization and Test-Time Sampling for Text-to-Image Generation

Gaussian Mixture Flow Matching Models

OmiAD: One-Step Adaptive Masked Diffusion Model for Multi-class Anomaly Detection via Adversarial Distillation

Continuous Visual Autoregressive Generation via Score Maximization

FrameBridge: Improving Image-to-Video Generation with Bridge Models

Distillation of Discrete Diffusion through Dimensional Correlations

Simple and Critical Iterative Denoising: A Recasting of Discrete Diffusion in Graph Generation

Scalable Non-Equivariant 3D Molecule Generation via Rotational Alignment

Differentiable Solver Search for Fast Diffusion Sampling

On the Guidance of Flow Matching

Effective and Efficient Masked Image Generation Models

Is Noise Conditioning Necessary for Denoising Generative Models?

Towards Rationale-Answer Alignment of LVLMs via Self-Rationale Calibration

ExLM: Rethinking the Impact of $\texttt{[MASK]}$ Tokens in Masked Language Models

Physics-informed Temporal Alignment for Auto-regressive PDE Foundation Models

M+: Extending MemoryLLM with Scalable Long-Term Memory

Fundamental Limits of Visual Autoregressive Transformers: Universal Approximation Abilities

Sparse Spectral Training and Inference on Euclidean and Hyperbolic Neural Networks

Oscillation-Reduced MXFP4 Training for Vision Transformers

Sorbet: A Neuromorphic Hardware-Compatible Transformer-Based Spiking Language Model

SageAttention2: Efficient Attention with Thorough Outlier Smoothing and Per-thread INT4 Quantization

Hgformer: Hyperbolic Graph Transformer for Collaborative Filtering

RePaViT: Scalable Vision Transformer Acceleration via Structural Reparameterization on Feedforward Network Layers

In-Context Denoising with One-Layer Transformers: Connections between Attention and Associative Memory Retrieval

In-Context Linear Regression Demystified: Training Dynamics and Mechanistic Interpretability of Multi-Head Softmax Attention

Contrastive Localized Language-Image Pre-Training

FG-CLIP: Fine-Grained Visual and Textual Alignment

QMamba: On First Exploration of Vision Mamba for Image Quality Assessment

Stream-level Flow Matching with Gaussian Processes

TabPFN Unleashed: A Scalable and Effective Solution to Tabular Classification Problems

GRADEO: Towards Human-Like Evaluation for Text-to-Video Generation via Multi-Step Reasoning

One-dimensional Path Convolution

Understanding Mode Connectivity via Parameter Space Symmetry

Multiobjective distribution matching

Compositional Scene Understanding through Inverse Generative Modeling

Explanatory Instructions: Towards Unified Vision Tasks Understanding and Zero-shot Generalization

Activation by Interval-wise Dropout: A Simple Way to Prevent Neural Networks from Plasticity Loss

Where is the Truth? The Risk of Getting Confounded in a Continual World

How Effective Can Dropout Be in Multiple Instance Learning ?

Larger or Smaller Reward Margins to Select Preferences for LLM Alignment?

Learning Bayesian Nash Equilibrium in Auction Games via Approximate Best Response

Position: LLMs Need a Bayesian Meta-Reasoning Framework for More Robust and Generalizable Reasoning

Position: Don't Use the CLT in LLM Evals With Fewer Than a Few Hundred Datapoints

Position: You Can't Manufacture a NeRF

Position: Solve Layerwise Linear Models First to Understand Neural Dynamical Phenomena (Neural Collapse, Emergence, Lazy/Rich Regime, and Grokking)

Suitability Filter: A Statistical Framework for Classifier Evaluation in Real-World Deployment Settings

De-AntiFake: Rethinking the Protective Perturbations Against Voice Cloning Attacks

Circumventing Backdoor Space via Weight Symmetry

Assessing Safety Risks and Quantization-aware Safety Patching for Quantized Large Language Models

TRUST-VLM: Thorough Red-Teaming for Uncovering Safety Threats in Vision-Language Models

Efficient and Privacy-Preserving Soft Prompt Transfer for LLMs

Improving Out-of-Distribution Detection via Dynamic Covariance Calibration

An End-to-End Model for Logits-Based Large Language Models Watermarking

Stealix: Model Stealing via Prompt Evolution

AdvPrompter: Fast Adaptive Adversarial Prompting for LLMs

Statistical Collusion by Collectives on Learning Platforms

Provably Cost-Sensitive Adversarial Defense via Randomized Smoothing

Understanding and Mitigating Miscalibration in Prompt Tuning for Vision-Language Models

Understanding Model Ensemble in Transferable Adversarial Attack

Scaling Laws for Differentially Private Language Models

Variance as a Catalyst: Efficient and Transferable Semantic Erasure Adversarial Attack for Customized Diffusion Models

Janus: Dual-Server Multi-Round Secure Aggregation with Verifiability for Federated Learning

The Value of Prediction in Identifying the Worst-Off

Noisy SIGNSGD Is More Differentially Private Than You (Might) Think

SecEmb: Sparsity-Aware Secure Federated Learning of On-Device Recommender System with Large Embedding

FOCoOp: Enhancing Out-of-Distribution Robustness in Federated Prompt Learning for Vision-Language Models

Fast Exact Unlearning for In-Context Learning Data for LLMs

Adapting to Linear Separable Subsets with Large-Margin in Differentially Private Learning

Towards Trustworthy Federated Learning with Untrusted Participants

Kandinsky Conformal Prediction: Beyond Class- and Covariate-Conditional Coverage

Self-Consuming Generative Models with Adversarially Curated Data

MONA: Myopic Optimization with Non-myopic Approval Can Mitigate Multi-step Reward Hacking

A New Approach to Backtracking Counterfactual Explanations: A Unified Causal Framework for Efficient Model Interpretability

A Lens into Interpretable Transformer Mistakes via Semantic Dependency

Archetypal SAE: Adaptive and Stable Dictionary Learning for Concept Extraction in Large Vision Models

The Illusion of Role Separation: Hidden Shortcuts in LLM Role Learning (and How to Fix Them)

(ends 7:00 PM)

7 p.m.

FRI 18 JUL

7:30 a.m.

Registration West

(ends 4:00 PM)

Registration East

(ends 12:00 PM)

8:20 a.m.

Workshop:

Programmatic Representations for Agent Learning

(ends 5:00 PM)

8:30 a.m.

Workshop:

Tiny Titans: The next wave of On-Device Learning for Foundation Models (TTODLer-FM)

(ends 5:15 PM)

Workshop:

2nd Generative AI for Biology Workshop

(ends 6:00 PM)

Workshop:

2nd AI for Math Workshop @ ICML 2025

(ends 6:00 PM)

Workshop:

The 1st Workshop on Vector Databases

(ends 5:30 PM)

8:45 a.m.

Workshop:

Assessing World Models: Methods and Metrics for Evaluating Understanding

(ends 5:15 PM)

Workshop:

ICML 2025 Workshop on Computational Optimization of Buildings (CO-BUILD)

(ends 6:00 PM)

Workshop:

Tokenization Workshop (TokShop)

(ends 5:30 PM)

9 a.m.

Break:

Coffee Break

(ends 10:00 AM)

Workshop:

3rd Workshop on High-dimensional Learning Dynamics (HiLD)

(ends 5:30 PM)

Workshop:

2nd Workshop on Models of Human Feedback for AI Alignment (MoFA)

(ends 5:00 PM)

Workshop:

Multi-Agent Systems in the Era of Foundation Models: Opportunities, Challenges and Futures

(ends 5:00 PM)

Workshop:

Machine Unlearning for Generative AI

(ends 5:00 PM)

Workshop:

Scaling Up Intervention Models

(ends 5:20 PM)

Workshop:

CODEML: Championing Open-source DEvelopment in Machine Learning

(ends 5:00 PM)

Workshop:

2nd Workshop on Test-Time Adaptation: Putting Updates to the Test (PUT)

(ends 5:50 PM)

Workshop:

1st Workshop on Foundation Models for Structured Data (FMSD)

(ends 5:00 PM)

9:15 a.m.

Workshop:

Machine Learning for Wireless Communication and Networks (ML4Wireless)

(ends 6:00 PM)

noon

Break:

Hosted Grab and Go Lunch

(ends 1:00 PM)

3 p.m.

Break:

Coffee Break

(ends 3:30 PM)

SAT 19 JUL

7:30 a.m.

Registration West

(ends 12:00 PM)

8 a.m.

Workshop:

Actionable Interpretability

(ends 5:00 PM)

8:25 a.m.

Workshop:

The Impact of Memorization on Trustworthy Foundation Models

(ends 5:00 PM)

8:30 a.m.

Workshop:

Exploration in AI Today (EXAIT)

(ends 5:15 PM)

Workshop:

Workshop on Multi-modal Foundation Models and Large Language Models for Life Sciences

(ends 5:40 PM)

Workshop:

Workshop on Computer Use Agents

(ends 6:00 PM)

Workshop:

ICML 2025 Workshop on Collaborative and Federated Agentic Workflows (CFAgentic @ ICML'25)

(ends 5:20 PM)

Workshop:

ES-FoMo III: 3rd Workshop on Efficient Systems for Foundation Models

(ends 5:10 PM)

8:45 a.m.

Workshop:

Methods and Opportunities at Small Scale (MOSS)

(ends 6:00 PM)

8:50 a.m.

Workshop:

The 2nd Workshop on Reliable and Responsible Foundation Models

(ends 5:05 PM)

Workshop:

Building Physically Plausible World Models

(ends 5:10 PM)

8:55 a.m.

Workshop:

DIG-BUGS: Data in Generative Models (The Bad, the Ugly, and the Greats)

(ends 5:00 PM)

Workshop:

DataWorld: Unifying data curation frameworks across domains

(ends 5:00 PM)

9 a.m.

Break:

Coffee Break

(ends 10:00 AM)

Workshop:

TerraBytes: Towards global datasets and models for Earth Observation

(ends 5:30 PM)

Workshop:

Workshop on Technical AI Governance

(ends 5:00 PM)

Workshop:

AI Heard That! ICML 2025 Workshop on Machine Learning for Audio

(ends 5:00 PM)

9:10 a.m.

Workshop:

The Second Workshop on Long-Context Foundation Models

(ends 7:00 PM)

noon

Break:

Hosted Grab and Go Lunch

(ends 1:00 PM)

3 p.m.

Break:

Coffee Break

(ends 3:30 PM)