Papers

12094 papers

ICLR2025

Bridging the Gap between Variational Inference and Stochastic Gradient MCMC in Function Space

Summary pending...

Bayesian deep learningfunctional variational inferencefunctional MCMC
ICLR2025

uniINF: Best-of-Both-Worlds Algorithm for Parameter-Free Heavy-Tailed MABs

Summary pending...

Heavy TailedMulti-Armed BanditsParameter-Free
ICLR2025

Q-Adapter: Customizing Pre-trained LLMs to New Preferences with Forgetting Mitigation

Summary pending...

reinforcement learning from human feedbacklarge languge model customizationanti-forgetting
ICLR2025

Adversarial Score identity Distillation: Rapidly Surpassing the Teacher in One Step

Summary pending...

Diffusion distillationdeep generative models
ICLR2025

Exploring the Design Space of Visual Context Representation in Video MLLMs

Summary pending...

Video Multimodal Large Language ModelScaling Law
ICLR2025

GeSubNet: Gene Interaction Inference for Disease Subtype Network Generation

Summary pending...

Gene Functional NetworksDisease SubtypesBioinformatics
ICLR2025

Shallow diffusion networks provably learn hidden low-dimensional structure

Summary pending...

Generative modelsdiffusion modelsdenoising score matching
ICLR2025

Iterative Label Refinement Matters More than Preference Optimization under Weak Supervision

Summary pending...

unreliable human supervisionlanguage model post-trainingscalable oversight
ICLR2025

CBQ: Cross-Block Quantization for Large Language Models

Summary pending...

Large Language Model Compressionultra-low bits precision
ICLR2025

Neural Causal Graph for Interpretable and Intervenable Classification

Summary pending...

Causal InferenceGraph Neural NetworkClassification
ICLR2025

ConvCodeWorld: Benchmarking Conversational Code Generation in Reproducible Feedback Environments

Summary pending...

Large language modelsMulti-turn code generationBenchmark
ICLR2025

Solving Token Gradient Conflict in Mixture-of-Experts for Large Vision-Language Model

Summary pending...

Large Vision-Language Model (LVLM)Mixture-of-Expert (MoE)token-level gradient
ICLR2025

Infilling Score: A Pretraining Data Detection Algorithm for Large Language Models

Summary pending...

Pretraining data detectionLarge language models
ICLR2025

StructRAG: Boosting Knowledge Intensive Reasoning of LLMs via Inference-time Hybrid Information Structurization

Summary pending...

RAGStructured Knowledge
ICLR2025

Spectral Compressive Imaging via Unmixing-driven Subspace Diffusion Refinement

Summary pending...

Spectral compressive imagingsubspacediffusion
ICLR2025

Tree-Wasserstein Distance for High Dimensional Data with a Latent Feature Hierarchy

Summary pending...

diffusion geometryhyperbolic geometrytree-Wasserstein distance
ICLR2025

Steering Protein Family Design through Profile Bayesian Flow

Summary pending...

protein family generationhomologous protein generationprotein design
ICLR2025

Looking Backward: Retrospective Backward Synthesis for Goal-Conditioned GFlowNets

Summary pending...

Backward SamplingGoal-Conditioned LearningGenerative Flow Networks
ICLR2025

RMP-SAM: Towards Real-Time Multi-Purpose Segment Anything

Summary pending...

segment anything; real-time segmentation; multi-purpose model;
ICLR2025

Emerging Safety Attack and Defense in Federated Instruction Tuning of Large Language Models

Summary pending...

Federated LearningLarge Language ModelsSafety Alignment