Yue Zhao

Assistant Professor@University of Southern California·美国·OpenReview

研究方向

Agent Safety · LLM Safety · AI Safety · Graph Neural Networks · out-of-distribution detection · automated machine learning · machine learning systems · anomaly detection · outlier detection

Yue Zhao

Doxing via the Lens: Revealing Location-related Privacy Leakage on Multi-modal Large Reasoning Models

TRIDENT: An Efficient Data-Free Model Extraction Attack for Graph Neural Networks

DecAlign: Hierarchical Cross-Modal Alignment for Decoupled Multimodal Representation Learning

Charts Are Not Images: On the Challenges of Scientific Chart Editing

ClimateLLM: Efficient Weather Forecasting via Frequency-Aware Large Language Models

TrustGen: A Platform of Dynamic Benchmarking on the Trustworthiness of Generative Foundation Models

Can Molecular Foundation Models Know What They Don't Know? A Simple Remedy with Preference Optimization

AutoDavis: Automatic and Dynamic Evaluation Protocol of Large Vision-Language Models on Visual Question-Answering

SocialMaze: A Benchmark for Evaluating Social Reasoning in Large Language Models

Selecting Out-of-Distribution Detector for Multiple Modalities

A Personalized Conversational Benchmark: Towards Simulating Personalized Conversations

Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs

Graph Synthetic Out-of-Distribution Exposure with Large Language Models

PCA Feature Alignment is Sufficient for Building Graph Foundation Models

DyFlow: Dynamic Workflow Framework for Agentic Reasoning

Meta-Learning for Speeding Up Large Model Inference in Decentralized Environments

JailDAM: Jailbreak Detection with Adaptive Memory for Vision-Language Model

MetaOOD: Automatic Selection of OOD Detection Models

Retrieval-Reasoning Large Language Model-based Synthetic Clinical Trial Generation