Previous Next

I’m working at Tencent AI Lab as a senior researcher. I received my Ph.D degree from the Chinese University of Hong Kong in 2021, under the supervision of Prof. Irwin King and Prof. Michael R. Lyu. Before that, I received my Bachelor degree and Mphil degree at Nanjing University in 2015 and 2017, respectively.

:office: Experiences

• Tencent AI Lab (2021 - Present), Senior Researcher
• Tencent AI Lab (2019 - 2021), Research Intern

:books: Research

Currently, I focus on Large Language Models (LLMs), including various interesting topics.

LLM Proactivity

1. Towards Evaluating Proactive Risk Awareness of Multimodal Language Models. Under Review 2025.
2. Learning to Ask: When LLM Agents Meet Unclear Instruction. EMNLP 2025.

LLM Continual Learning

1. REA-RL: Reflection-Aware Online Reinforcement Learning for Efficient Large Reasoning Models. Under Review 2025.
2. DRPruning: Efficient Large Language Model Pruning through Distributionally Robust Optimization. ACL 2025.
3. NewTerm: Benchmarking Real-Time New Terms for Large Language Models with Annual Updates. NeurIPS 2024 D&B.
4. Tencent’s Multilingual Machine Translation System for WMT22 Large-Scale African Languages. WMT 2022.

LLM Agents & Reasoning

1. How Far Are We on the Decision-Making of LLMs? Evaluating LLMs’ Gaming Ability in Multi-Agent Environments. ICLR 2025.
2. LogicAsker: Evaluating and Improving the Logical Reasoning Ability of Large Language Models. EMNLP 2024.
3. Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate. EMNLP 2024.
4. CoAct: A Global-Local Hierarchy for Autonomous Agent Collaboration. Preprint 2024.

LLM Safety

1. Refuse Whenever You Feel Unsafe: Improving Safety in LLMs via Decoupled Refusal Training. ACL 2025.
2. Chain-of-Jailbreak Attack for Image Generation Models via Step by Step Editing. Findings of ACL 2025.
3. GPT-4 Is Too Smart To Be Safe: Stealthy Chat with LLMs via Cipher. ICLR 2024.
4. Not all countries celebrate thanksgiving: On the cultural dominance in large language models. ACL 2024.
5. All languages matter: On the multilingual safety of large language models. Findings of ACL 2024.

LLM Personality

1. Emotionally Numb or Empathetic? Evaluating How LLMs Feel using EmotionBench. NeurIPS 2024.
2. On the reliability of psychological scales on large language models. EMNLP 2024.
3. On the Humanity of Conversational AI: Evaluating the Psychological Portrayal of LLMs. ICLR 2024 Oral.

Sign Language Translation

1. Improving Gloss-free Sign Language Translation by Reducing Representation Density. NeurIPS 2024.
2. Unsupervised Sign Language Translation and Generation. Findings of ACL 2024.
3. Cross-modality Data Augmentation for End-to-End Sign Language Translation. Findings of EMNLP 2023.

:fire: Spotlight Projects

• : LLM-Hands-On
• : Is ChatGPT A Good Translator
• : Translating during Chat: ParroT
• : Human-Like Translation Strategy
• : Multi-Agent Debate
• : CipherChat
• : PsychoBench