Zhenkai Liang

AlphaSteer: Learning Refusal Steering with Principled Null-Space Constraint

DevOps-Gym: Benchmarking AI Agents in Software DevOps Cycle

AlphaAlign: Incentivizing Safety Alignment with Extremely Simplified Reinforcement Learning

OSS-Bench: Evaluating LLMs via the Realistic Open-Source Software Development Pipeline

Equipping LLMs with Self-Awareness for High-Stakes Tasks

AttackSeqBench: Benchmarking Large Language Models in Analyzing Attack Sequences within Cyber Threat Intelligence

RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards