Papers
12094 papers
ICML2025
On Mitigating Affinity Bias through Bandits with Evolving Biased Feedback
Summary pending...
FairnessBanditsLower Bounds
ICML2025
Do We Need to Verify Step by Step? Rethinking Process Supervision from a Theoretical Perspective
Summary pending...
Reinforcement Learning TheoryProcess SupervisionOutcome Supervision
ICML2025
Benefits of Early Stopping in Gradient Descent for Overparameterized Logistic Regression
Summary pending...
Implicit RegularizationGDEarly Stopping
ICML2025
A General Representation-Based Approach to Multi-Source Domain Adaptation
Summary pending...
Domain adaptationrepresentation learning
ICML2025
How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias
Summary pending...
TransformersTraining dynamicsImplicit bias
ICML2025
Compositional Causal Reasoning Evaluation in Language Models
Summary pending...
language modelscausal reasoningcompositional reasoning
ICML2025
Generalists vs. Specialists: Evaluating LLMs on Highly-Constrained Biophysical Sequence Optimization Tasks
Summary pending...
biophysical optimizationlarge language modelsdiscrete sequence optimization
ICML2025
Mahalanobis++: Improving OOD Detection via Feature Normalization
Summary pending...
OOD detectionMahalanobis distanceout-of-distribution detection
ICML2025
Aligning Spoken Dialogue Models from User Interactions
Summary pending...
Speech AlignmentAudio Language ModelConversational Model
ICML2025
Selective Preference Aggregation
Summary pending...
RankingDisagreementPreference Aggregation
ICML2025
A Unified Approach to Routing and Cascading for LLMs
Summary pending...
large language modelsroutingcascading
ICML2025
Provable Zero-Shot Generalization in Offline Reinforcement Learning
Summary pending...
offline reinforcement learninggeneralization
ICML2025
Generalization Analysis for Controllable Learning
Summary pending...
ControllabilityGeneralization Analysis
ICML2025
QuEst: Enhancing Estimates of Quantile-Based Distributional Measures Using Model Predictions
Summary pending...
prediction-powered inferenceL-statisticsauto-evaluation
ICML2025
Reinforce LLM Reasoning through Multi-Agent Reflection
Summary pending...
Post-trainingLLM-based multi-agentsReinforcement learning
ICML2025
Verification Learning: Make Unsupervised Neuro-Symbolic System Feasible
Summary pending...
Neural Symbolic Learning
ICML2025
From Language Models over Tokens to Language Models over Characters
Summary pending...
tokenizationcharacterbytes
ICML2025
Stacey: Promoting Stochastic Steepest Descent via Accelerated $\ell_p$-Smooth Nonconvex Optimization
Summary pending...
Non-convex OptimizationNon-Euclidean AccelerationStochastic Steepest Descent
ICML2025
Diffusion on Language Model Encodings for Protein Sequence Generation
Summary pending...
continuous diffusiongenerative protein designscore matching
ICML2025
Layer by Layer: Uncovering Hidden Representations in Language Models
Summary pending...
large language modelentropyaugmentation