Papers

12094 papers

ICML2025

The Role of Randomness in Stability

Summary pending...

ReplicabilityDifferential PrivacyPAC Learning
ICML2025

Causal Abstraction Inference under Lossy Representations

Summary pending...

causalitycausal inferencecausal abstractions
ICML2025

(How) Can Transformers Predict Pseudo-Random Numbers?

Summary pending...

InterpretabilityIn-context learningGrokking
ICML2025

ConceptAttention: Diffusion Transformers Learn Highly Interpretable Features

Summary pending...

diffusioninterpretabilitytransformers
ICML2025

Uncertainty Quantification for LLM-Based Survey Simulations

Summary pending...

synthetic datalarge language modelsuncertainty quantification
ICML2025

Geometry Informed Tokenization of Molecules for Language Model Generation

Summary pending...

Language models
ICML2025

Floating-Point Neural Networks Can Represent Almost All Floating-Point Functions

Summary pending...

Universal approximationFloating-Point Arithmetic
ICML2025

Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worse

Summary pending...

chain of thoughtpsychologycognitive science
ICML2025

Tracking Most Significant Shifts in Infinite-Armed Bandits

Summary pending...

non-stationaryinfinite-armed banditsbandits
ICML2025

ELoRA: Low-Rank Adaptation for Equivariant GNNs

Summary pending...

graph neural networkequivarianceparameter-efficient fine-tuning
ICML2025

Dynamical Modeling of Behaviorally Relevant Spatiotemporal Patterns in Neural Imaging Data

Summary pending...

Deep LearningDynamical ModelingNeural Imaging
ICML2025

Test-Time Training Provably Improves Transformers as In-context Learners

Summary pending...

in-context learningtransformerstest-time training
ICML2025

Windows Agent Arena: Evaluating Multi-Modal OS Agents at Scale

Summary pending...

agentsbenchmarkcomputer agents
ICML2025

Graph World Model

Summary pending...

Graph world modelUnstructured dataMulti-modal information
ICML2025

LLM-SRBench: A New Benchmark for Scientific Equation Discovery with Large Language Models

Summary pending...

BenchmarkScientific DiscoveryLarge Language Models
ICML2025

Fast, Accurate Manifold Denoising by Tunneling Riemannian Optimization

Summary pending...

Manifold DenoisingLearning-to-optimize
ICML2025

Federated In-Context Learning: Iterative Refinement for Improved Answer Quality

Summary pending...

Federated Learning; In-context Learning; Large Language Model; Natural Language Processing
ICML2025

Triple-Optimistic Learning for Stochastic Contextual Bandits with General Constraints

Summary pending...

tripe-optimitistic frameworkcontetxual banditsgeneral constraints
ICML2025

In-Context Linear Regression Demystified: Training Dynamics and Mechanistic Interpretability of Multi-Head Softmax Attention

Summary pending...

Multi-head AttentionMechanism IntepretationTraining Dynamics
ICML2025

RZ-NAS: Enhancing LLM-guided Neural Architecture Search via Reflective Zero-Cost Strategy

Summary pending...

Automated Machine LearningNeural Architecture SearchLarge Language Models