Yujun Zhou

Building a Foundational Guardrail for General Agentic Systems via Synthetic Data

Causally-Enhanced Reinforcement Policy Optimization of Large Language Models

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

Emergent Deceptive Behaviors in Reward-Optimizing LLMs

On the Evolution of Language Models without Labels: Majority Drives Selection, Novelty Promotes Variation

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

CLUE: Non-parametric Verification from Experience via Hidden-State Clustering

AdaReasoner: Adaptive Reasoning Enables More Flexible Thinking

Exposing and Patching the Flaws of Large Language Models in Social Character Simulation

ChemOrch: Empowering LLMs with Chemical Intelligence via Groundbreaking Synthetic Instructions

Benchmarking LLMs on Safety Issues in Scientific Labs