Hongyang Zhang

20

Papers

181

Total Citations

Papers (20)

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

Noise-Tolerant Interactive Learning Using Pairwise Comparisons

Noise-Tolerant Life-Long Matrix Completion via Adaptive Sampling

NeurIPS 2016arXiv

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

Sample and Computationally Efficient Learning Algorithms under S-Concave Distributions

NeurIPS 2017arXiv

EAGLE: Speculative Sampling Requires Rethinking Feature Uncertainty

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

Design and Interpretation of Universal Adversarial Patches in Face Detection

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

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

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

Efficient Symmetric Norm Regression via Linear Sketching

Optimal Analysis of Subset-Selection Based L_p Low-Rank Approximation

A Closer Look at Accuracy vs. Robustness

Self-Adaptive Training: beyond Empirical Risk Minimization

Improved Regularization and Robustness for Fine-tuning in Neural Networks

Boosting Barely Robust Learners: A New Perspective on Adversarial Robustness

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

Theoretically Principled Trade-off between Robustness and Accuracy

Differentially Private Clustering in High-Dimensional Euclidean Spaces