Papers
12094 papers
ICML2025
Is Best-of-N the Best of Them? Coverage, Scaling, and Optimality in Inference-Time Alignment
Summary pending...
inference time alignmentbest of npessimism
ICML2025
Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration
Summary pending...
Cooperative Multi-Agent Reinforcement LearningState ModellingSparse Reward
ICML2025
TANGO: Clustering with Typicality-Aware Nonlocal Mode-Seeking and Graph-Cut Optimization
Summary pending...
clusteringdensity-based clusteringmode-seeking
ICML2025
CALM: Consensus-Aware Localized Merging for Multi-Task Learning
Summary pending...
model merging; multi-task learning; global task consensus; localized merging
ICML2025
SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning
Summary pending...
Language ModelsAlignmentPreference Optimization
ICML2025
DRAG: Data Reconstruction Attack using Guided Diffusion
Summary pending...
Data Reconstruction AttackPrivacyDiffusion Model
ICML2025
VCT: Training Consistency Models with Variational Noise Coupling
Summary pending...
Consistency ModelsGenerative Models
ICML2025
The Missing Alignment Link of In-context Learning on Sequences
Summary pending...
Model adaptationfew-shot learningsequence to sequence learning
ICML2025
Sparse Autoencoders, Again?
Summary pending...
autoencoderssparse representationsvariational autoencoders
ICML2025
Unveiling Markov heads in Pretrained Language Models for Offline Reinforcement Learning
Summary pending...
decision transformer; cross-domain
ICML2025
Integrating Intermediate Layer Optimization and Projected Gradient Descent for Solving Inverse Problems with Diffusion Models
Summary pending...
diffusion modelinverse problems
ICML2025
Latent Variable Causal Discovery under Selection Bias
Summary pending...
causal discoverylatent variablesselection bias
ICML2025
Learning Efficient Robotic Garment Manipulation with Standardization
Summary pending...
Deformable Object Manipulation，Bimanual Manipulation，Self-supervised learning
ICML2025
The Berkeley Function Calling Leaderboard (BFCL): From Tool Use to Agentic Evaluation of Large Language Models
Summary pending...
Function Calling EvaluationTool useAgentic Evaluation
ICML2025
Deep Fuzzy Multi-view Learning for Reliable Classification
Summary pending...
Multi-view LearningTrustworthy Machine LearningFuzzy Deep Learning
ICML2025
LipsNet++: Unifying Filter and Controller into a Policy Network
Summary pending...
Deep Reinforcement LearningPolicy Network DesignAction Fluctuation
ICML2025
BOPO: Neural Combinatorial Optimization via Best-anchored and Objective-guided Preference Optimization
Summary pending...
Neural Combinatorial Optimization; Preference Optimization; Machine Learning;
ICML2025
A Memory Efficient Randomized Subspace Optimization Method for Training Large Language Models
Summary pending...
randomized subspace optimizationlarge language model trainingstochastic optimization
ICML2025
Graph-Supported Dynamic Algorithm Configuration for Multi-Objective Combinatorial Optimization
Summary pending...
Deep Reinforcement LearningMulti-objective Combinatorial OptimizationDynamic Algorithm Configuration
ICML2025
Alpha-SQL: Zero-Shot Text-to-SQL using Monte Carlo Tree Search
Summary pending...
Text-to-SQLDatabasesLLMs