Jieyu Zhang

Affiliations

University of Illinois Urbana-ChampaignUniversity of Washington

papers

1,491

total citations

papers (16)

DataComp: In search of the next generation of multimodal datasets

NEURIPS 2023arXiv

608

citations

Large Language Model as Attributed Training Data Generator: A Tale of Diversity and Bias

NEURIPS 2023arXiv

324

citations

SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality

NEURIPS 2023arXiv

195

citations

SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models

ICML 2024arXiv

181

citations

Frustratingly Easy Regularization on Representation Can Boost Deep Reinforcement Learning

CVPR 2023arXiv

citations

Discovering Knowledge Deficiencies of Language Models on Massive Knowledge Base

COLM 2025arXiv

citations

One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory

ICCV 2025arXiv

citations

Characterizing the Impacts of Semi-supervised Learning for Weak Supervision

NEURIPS 2023

citations

Optimizing Information-theoretical Generalization Bound via Anisotropic Noise of SGLD

NEURIPS 2021

citations

Jieyu Zhang

Affiliations

papers (16)

DataComp: In search of the next generation of multimodal datasets

Large Language Model as Attributed Training Data Generator: A Tale of Diversity and Bias

SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality

SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models

Subclass-balancing Contrastive Learning for Long-tailed Recognition

Offline Training of Language Model Agents with Functions as Learnable Weights

When to Learn What: Model-Adaptive Data Augmentation Curriculum

Uncovering Neural Scaling Laws in Molecular Representation Learning

Iterated Learning Improves Compositionality in Large Vision-Language Models

Understanding Programmatic Weak Supervision via Source-aware Influence Function

On the Trade-off of Intra-/Inter-class Diversity for Supervised Pre-training

Frustratingly Easy Regularization on Representation Can Boost Deep Reinforcement Learning

Discovering Knowledge Deficiencies of Language Models on Massive Knowledge Base

One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory

Characterizing the Impacts of Semi-supervised Learning for Weak Supervision

Optimizing Information-theoretical Generalization Bound via Anisotropic Noise of SGLD

papers (16)

DataComp: In search of the next generation of multimodal datasets

Large Language Model as Attributed Training Data Generator: A Tale of Diversity and Bias

SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality

SciBench: Evaluating College-Level Scientific Problem-Solving Abilities of Large Language Models

Subclass-balancing Contrastive Learning for Long-tailed Recognition

Offline Training of Language Model Agents with Functions as Learnable Weights

When to Learn What: Model-Adaptive Data Augmentation Curriculum

Uncovering Neural Scaling Laws in Molecular Representation Learning

Iterated Learning Improves Compositionality in Large Vision-Language Models

Understanding Programmatic Weak Supervision via Source-aware Influence Function

On the Trade-off of Intra-/Inter-class Diversity for Supervised Pre-training

Frustratingly Easy Regularization on Representation Can Boost Deep Reinforcement Learning

Discovering Knowledge Deficiencies of Language Models on Massive Knowledge Base

One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory

Characterizing the Impacts of Semi-supervised Learning for Weak Supervision

Optimizing Information-theoretical Generalization Bound via Anisotropic Noise of SGLD