Hongyang Zhang

Affiliations

University of WaterlooVector Institute for AI

papers

1,159

total citations

papers (14)

EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test

NEURIPS 2025arXiv

115

citations

AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API Calls

ICML 2024arXiv

citations

Improved Regularization and Robustness for Fine-tuning in Neural Networks

NEURIPS 2021arXiv

citations

Design and Interpretation of Universal Adversarial Patches in Face Detection

ECCV 2020arXiv

citations

A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models

ICML 2024arXiv

citations

A Closer Look at Accuracy vs. Robustness

NEURIPS 2020arXiv

citations

Graph Neural Networks for Road Safety Modeling: Datasets and Evaluations for Accident Analysis

NEURIPS 2023arXiv

citations

Lost Domain Generalization Is a Natural Consequence of Lack of Training Domains

AAAI 2024

citations

BACON: Improving Clarity of Image Captions via Bag-of-Concept Graphs

CVPR 2025arXiv

citations

Boosting Barely Robust Learners: A New Perspective on Adversarial Robustness

NEURIPS 2022arXiv

citations

Cooperation or Competition: Avoiding Player Domination for Multi-Target Robustness via Adaptive Budgets

CVPR 2023arXiv

citations

Hongyang Zhang

Affiliations

papers (14)

EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty

Self-Adaptive Training: beyond Empirical Risk Minimization

RAIN: Your Language Models Can Align Themselves without Finetuning

EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test

AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API Calls

Improved Regularization and Robustness for Fine-tuning in Neural Networks

Design and Interpretation of Universal Adversarial Patches in Face Detection

A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models

A Closer Look at Accuracy vs. Robustness

Graph Neural Networks for Road Safety Modeling: Datasets and Evaluations for Accident Analysis

Lost Domain Generalization Is a Natural Consequence of Lack of Training Domains

BACON: Improving Clarity of Image Captions via Bag-of-Concept Graphs

Boosting Barely Robust Learners: A New Perspective on Adversarial Robustness

Cooperation or Competition: Avoiding Player Domination for Multi-Target Robustness via Adaptive Budgets

papers (14)

EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty

Self-Adaptive Training: beyond Empirical Risk Minimization

RAIN: Your Language Models Can Align Themselves without Finetuning

EAGLE-3: Scaling up Inference Acceleration of Large Language Models via Training-Time Test

AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API Calls

Improved Regularization and Robustness for Fine-tuning in Neural Networks

Design and Interpretation of Universal Adversarial Patches in Face Detection

A Resilient and Accessible Distribution-Preserving Watermark for Large Language Models

A Closer Look at Accuracy vs. Robustness

Graph Neural Networks for Road Safety Modeling: Datasets and Evaluations for Accident Analysis

Lost Domain Generalization Is a Natural Consequence of Lack of Training Domains

BACON: Improving Clarity of Image Captions via Bag-of-Concept Graphs

Boosting Barely Robust Learners: A New Perspective on Adversarial Robustness

Cooperation or Competition: Avoiding Player Domination for Multi-Target Robustness via Adaptive Budgets