|
Wei-Lin Chen I am a 2nd-year CS PhD student at University of Virginia and a part-time student researcher at Google. I am fortunate to be advised by Yu Meng and am grateful to be supported by UVA Provost's Fellowship and UVA Computer Science Scholar Fellowship. Nowadays, I think about (1) how to measure LLM's reasoning effort beyond superficial features like token length, and (2) how to ensure LLM-based evaluators succeed in non-trivially verifiable domains by grounding their behavior in insights drawn from verifiable tasks. Google Scholar / LinkedIn / CV / X (Twitter) / Email |
|
Select Research*equal contribution |
|
Do LLM Evaluators Prefer Themselves for a Reason?
Wei-Lin Chen, Zhepei Wei, Xinyu Zhu, Shi Feng, Yu Meng Preprint arXiv / thread |
|
The Surprising Effectiveness of Negative Reinforcement in LLM Reasoning
Xinyu Zhu, Mengzhou Xia, Zhepei Wei, Wei-Lin Chen, Danqi Chen, Yu Meng NeurIPS 2025 arXiv / thread |
|
Evaluating Large Language Models as Expert Annotators
Yu-Min Tseng, Wei-Lin Chen, Chung-Chi Chen, Hsin-Hsi Chen COLM 2025 arXiv / thread |
|
InstructRAG: Instructing Retrieval-Augmented Generation via Self-Synthesized Rationales
Zhepei Wei, Wei-Lin Chen, Yu Meng ICLR 2025 arXiv / thread |
|
Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization
Yu-Min Tseng\(^*\), Yu-Chao Huang\(^*\), Teng-Yun Hsiao\(^*\),Wei-Lin Chen\(^*\), Chao-Wei Huang, Yu Meng, Yun-Nung Chen EMNLP 2024 Findings arXiv / repo |
|
Self-ICL: Zero-Shot In-Context Learning with Self-Generated Demonstrations
Wei-Lin Chen\(^*\), Cheng-Kuang Wu\(^*\), Yun-Nung Chen, Hsin-Hsi Chen EMNLP 2023 arXiv / poster |