Papers

12094 papers

ICML2025

Convergence of Consistency Model with Multistep Sampling under General Data Assumptions

Summary pending...

Consistency modelsdiffusion modelslearning theory
ICML2025

Robust Reward Alignment via Hypothesis Space Batch Cutting

Summary pending...

Learning from Human FeedbackInverse Reinforcement LearningPreference Based Reinforcement Learning
ICML2025

High-Dimensional Prediction for Sequential Decision Making

Summary pending...

online decision makingcombinatorial optimizationmulticalibration
ICML2025

Improving Model Alignment Through Collective Intelligence of Open-Source Models

Summary pending...

AlignmentOpen-Source ModelMixture of Agents
ICML2025

GSM-$\infty$: How Do your LLMs Behave over Infinitely Increasing Reasoning Complexity and Context Length?

Summary pending...

Long ContextReasoningUnderstanding
ICML2025

No Free Lunch from Random Feature Ensembles: Scaling Laws and Near-Optimality Conditions

Summary pending...

Ensemble LearningRegressionScaling Laws
ICML2025

What can large language models do for sustainable food?

Summary pending...

large language modelssustainabilityclimate
ICML2025

Do Vision-Language Models Really Understand Visual Language?

Summary pending...

vision-language modelvisual languageevaluation
ICML2025

Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflow

Summary pending...

Flow-matchingfew-shot generationequivariance
ICML2025

Training a Generally Curious Agent

Summary pending...

LLM AgentSynethic DataMultiturn finetuning
ICML2025

Putnam-AXIOM: A Functional & Static Benchmark for Measuring Higher Level Mathematical Reasoning in LLMs

Summary pending...

BenchmarksLarge Language ModelsMathematical Reasoning
ICML2025

Hardware and Software Platform Inference

Summary pending...

ML securityML governance
ICML2025

Value-Based Deep RL Scales Predictably

Summary pending...

scaling lawsonline reinforcement learningq-learning
ICML2025

DeepCrossAttention: Supercharging Transformer Residual Connections

Summary pending...

residual networkcross attentionresnet
ICML2025

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

Summary pending...

mathematical reasoningbenchmarkrobustness
ICML2025

Policy-Regret Minimization in Markov Games with Function Approximation

Summary pending...

policy regretMarkov gamesstrategic opponents
ICML2025

SWE-Lancer: Can Frontier LLMs Earn $1 Million from Real-World Freelance Software Engineering?

Summary pending...

software engineeringbenchmarkevals
ICML2025

Universal Neural Optimal Transport

Summary pending...

Optimal TransportNeural OperatorsMeta Learning
ICML2025

RLEF: Grounding Code LLMs in Execution Feedback with Reinforcement Learning

Summary pending...

large language modelsmulti-turn code generationreinforcement learning
ICML2025

PENCIL: Long Thoughts with Short Memory

Summary pending...

Large language modelschain-of-thought