Papers

12094 papers

ICML2025

From Low Rank Gradient Subspace Stabilization to Low-Rank Weights: Observations, Theories, and Applications

Summary pending...

Large language modelsGradient subspaceMemory-efficient training
ICML2025

On the Robustness of Reward Models for Language Model Alignment

Summary pending...

RLHFReward modelingOver-optimization
ICML2025

A Bregman Proximal Viewpoint on Neural Operators

Summary pending...

neural operatorsproximal optimizationbregman divergence
ICML2025

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

Summary pending...

image-text datasets; synthetic captions
ICML2025

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

Summary pending...

Analog circuitcircuit topologyreinforcement learning
ICML2025

Do Bayesian Neural Networks Actually Behave Like Bayesian Models?

Summary pending...

Bayesian Neural NetworksBNNsBayesian Deep Learning
ICML2025

Statistical Query Hardness of Multiclass Linear Classification with Random Classification Noise

Summary pending...

Multiclass Linear ClassificationRandom Classification NoiseStatistical Query Learning
ICML2025

Synthetic Text Generation for Training Large Language Models via Gradient Matching

Summary pending...

Synthetic dataLarge language modelsGradient matching
ICML2025

Preference Controllable Reinforcement Learning with Advanced Multi-Objective Optimization

Summary pending...

Multi-Objective OptimizationMulti-Objective Reinforcement Learning
ICML2025

SCISSOR: Mitigating Semantic Bias through Cluster-Aware Siamese Networks for Robust Classification

Summary pending...

Semantic BiasShortcut LearningSiamese Networks
ICML2025

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

Summary pending...

Foundation ModelIn-Context LearningTabular Data
ICML2025

Understanding Overadaptation in Supervised Fine-Tuning: The Role of Ensemble Methods

Summary pending...

model ensemblefine-tuningmulti-task
ICML2025

Zero-Shot Generalization of GNNs over Distinct Attribute Domains

Summary pending...

GNNzero-shotgraph foundation models
ICML2025

Randomized Dimensionality Reduction for Euclidean Maximization and Diversity Measures

Summary pending...

Dimensionality ReductionDimension ReductionGeometric Optimization
ICML2025

DIS-CO: Discovering Copyrighted Content in VLMs Training Data

Summary pending...

Copyrighted Content DetectionMembership Inference AttacksLarge Vision Language Models
ICML2025

Improving the Scaling Laws of Synthetic Data with Deliberate Practice

Summary pending...

Synthetic DataDeliberate PracticeActive Learning
ICML2025

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

Summary pending...

LLM AgentMulti-Agent System
ICML2025

RULEBREAKERS: Challenging LLMs at the Crossroads between Formal Logic and Human-like Reasoning

Summary pending...

reasoninglogichuman-like reasoning
ICML2025

Leveraging Online Olympiad-Level Math Problems for LLMs Training and Contamination-Resistant Evaluation

Summary pending...

Mathematical ReasoningLarge Language ModelsEvaluation
ICML2025

CRANE: Reasoning with constrained LLM generation

Summary pending...

LLM reasoningConstrained DecodingGrammar Guided Generation.