Papers

12094 papers

ICLR2024

PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback

Summary pending...

Reinforcement LearningPolicy optimizationPolicy alignment
ICLR2024

CAS: A Probability-Based Approach for Universal Condition Alignment Score

Summary pending...

Generative modeldiffusion modelscore-based prior
ICLR2024

Backdoor Contrastive Learning via Bi-level Trigger Optimization

Summary pending...

backdoor attackunsupervised contrastive learning
ICLR2024

How to Fine-Tune Vision Models with SGD

Summary pending...

fine-tuningSGDfreezing layers
ICLR2024

Evaluating Large Language Models at Evaluating Instruction Following

Summary pending...

large language modelsinstruction tuningevaluation
ICLR2024

Vanishing Gradients in Reinforcement Finetuning of Language Models

Summary pending...

Vanishing GradientsReinforcement FinetuningSupervised Finetuning
ICLR2024

Grounding Language Plans in Demonstrations Through Counterfactual Perturbations

Summary pending...

Grounding LLMLearning Mode Abstractions for ManipulationLearning from Demonstration
ICLR2024

Towards Understanding Factual Knowledge of Large Language Models

Summary pending...

Large Language ModelsResource and EvaluationInterpretability
ICLR2024

Contextual Bandits with Online Neural Regression

Summary pending...

Neural BanditsContextual BanditsRegret Bounds
ICLR2024

Bootstrapping Variational Information Pursuit with Large Language and Vision Models for Interpretable Image Classification

Summary pending...

Interpretable MLExplainable AIInformation Pursuit
ICLR2024

Plug-and-Play Policy Planner for Large Language Model Powered Dialogue Agents

Summary pending...

Dialogue Policy PlanningProactive DialogueLarge Language Model
ICLR2024

Kill Two Birds with One Stone: Rethinking Data Augmentation for Deep Long-tailed Learning

Summary pending...

Long-tailed learningdata augmentationimbalance
ICLR2024

In defense of parameter sharing for model-compression

Summary pending...

parameter sharingmodel compressionpruning
ICLR2024

Beating Price of Anarchy and Gradient Descent without Regret in Potential Games

Summary pending...

q-replicator dynamicspotential gamesaverage price of anarchy
ICLR2024

Understanding Addition in Transformers

Summary pending...

InterpretabilityTransformers
ICLR2024

Beyond IID weights: sparse and low-rank deep Neural Networks are also Gaussian Processes

Summary pending...

Deep Neural NetworksGaussian processesNeural Networks initialisation
ICLR2024

Harnessing Joint Rain-/Detail-aware Representations to Eliminate Intricate Rains

Summary pending...

Joint rain-/detail-aware representation learningcontrastive learningcontext-based modulation mechanism
ICLR2024

Differentially Private Synthetic Data via Foundation Model APIs 1: Images

Summary pending...

synthetic datadifferential privacymodel API
ICLR2024

Image Background Serves as Good Proxy for Out-of-distribution Data

Summary pending...

Out-of-distribution detectionOOD supervisionrobust image classification
ICLR2024

AgentBench: Evaluating LLMs as Agents

Summary pending...

Large language modelsAutonomous agentsReasoning