Papers

12094 papers

ICML2025

AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML

Summary pending...

Automated Machine LearningMulti-Agent FrameworkLarge Language Models
ICML2025

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

Summary pending...

DecoupleCross-DomainFew-Shot
ICML2025

Faster Approximation Algorithms for k-Center via Data Reduction

Summary pending...

k-centerclusteringEuclidean
ICML2025

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

Summary pending...

Federated LearningDifferential PrivacyMembership Inference Attack
ICML2025

TSP: A Two-Sided Smoothed Primal-Dual Method for Nonconvex Bilevel Optimization

Summary pending...

Primal-dual methodbilevel optimizationstochastic gradient descent
ICML2025

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

Summary pending...

Continual LearningIncremental Learning
ICML2025

Self-Consistency Preference Optimization

Summary pending...

Self-AlignmentUnsupervised and Semi-supervised ReasoningIterative Training
ICML2025

ProDiff: Prototype-Guided Diffusion for Minimal Information Trajectory Imputation

Summary pending...

Minimal Information Trajectory ImputationPrototype LearningDiffusion Model
ICML2025

Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning

Summary pending...

Mixture of Expertsfeature learningsingle-index models
ICML2025

Text-to-CAD Generation Through Infusing Visual Feedback in Large Language Models

Summary pending...

Large Language ModelsComputer-aided DesignText-to-CAD Generation
ICML2025

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

Summary pending...

DisentanglementCompositionFew-shot
ICML2025

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

Summary pending...

Sparse autoencodersmechanistic interpretabilityevaluations
ICML2025

Expressive Power of Graph Neural Networks for (Mixed-Integer) Quadratic Programs

Summary pending...

Graph Neural NetworksQuadratic ProgrammingLearning to Optimize
ICML2025

Highly Compressed Tokenizer Can Generate Without Training

Summary pending...

image tokenizer1D tokenizerautoencoder
ICML2025

ResearchTown: Simulator of Human Research Community

Summary pending...

Multi-agent; Graph Neural Network; Large Language Model; Automatic Research
ICML2025

Geometric Algebra Planes: Convex Implicit Neural Volumes

Summary pending...

Volume representationtensor decompositionconvex optimization
ICML2025

Efficient ANN-SNN Conversion with Error Compensation Learning

Summary pending...

ANN-SNN ConversionDeep Spiking Neural Networkbrain-inspired computing
ICML2025

Pruning for GNNs: Lower Complexity with Comparable Expressiveness

Summary pending...

Graph Neural NetworkPruningExpressiveness
ICML2025

Maximum Coverage in Turnstile Streams with Applications to Fingerprinting Measures

Summary pending...

maximum coveragelinear sketchesturnstile streams
ICML2025

SAFER: A Calibrated Risk-Aware Multimodal Recommendation Model for Dynamic Treatment Regimes

Summary pending...

Dynamic Treatment RegimeMultimodalRisk-aware