Yafu Li (李雅夫)

I am a researcher in Shanghai / Pujiang AI Lab, under the supervision of Prof. Yu Cheng. I earned my PhD through joint training at Zhejiang University and Westlake University under the guidance of Prof. Yue Zhang. During my internship at Tencent AI Lab, I collaborated closely with Dr. Leyang Cui and Dr. Wei Bi.

I earned my Bachelor’s degree at Wuhan University and subsequently pursued a Master’s degree at the University of Edinburgh, where I was supervised by Prof. Alex Lascarides. Prior to my PhD, I worked as an NLP researcher in Noah Ark's lab at Huawei, under the mentorship of Dr. Liangyou Li and Prof. Qun Liu.

Email  /  Google Scholar  /  Semantic Scholar  /  Twitter  /  Github

We are looking for interns and joint PhD candidates (with THU, PKU, SJTU, FDU, etc.) to work on cutting-edge research in large language models. Our focus areas include zero reinforcement learning (e.g., R1-zero), test-time scaling, and trustworthy AI. If you are interested, please feel free to contact me at yafuly@gmail.com.

My research focuses on test-time scaling, trustworthy AI and machine translation. * denotes equal contributions.

We survey recent efforts to improve reasoning efficiency in Large Reasoning Models, highlighting inefficiency patterns in long reasoning traces and discussing solutions across the model lifecycle from pretraining to inference.

We propose Mixture-of-Search-Agents (MoSA), a test-time framework that coordinates multiple LLMs via Monte Carlo Tree Search to improve complex reasoning by aggregating diverse search trajectories.

We present Test-Time Preference Optimization (TPO) that aligns LLMs during inference, surpassing strong baselines aligned during training.

We present Aggregation Fine-Tuning (AFT) where the model learns to aggregate multiple drafts into a single answer.

We present a comprehensive benchmark dataset designed to assess the proficiency of machine-generated text detectors amidst real-world scenarios.

We propose a novel task to identify "AI-touched" text spans in a fine-grained manner.

A survey of hallucination in LLMs.

We propose a syntax-guided generation schema, which generates the sequence guided by a constituency parse tree in a top-down direction.

We propose multi-granularity optimization for non-autoregressive translation, which collects model behaviors on translation segments of various granularities and integrates feedback for backpropagation.

We present a systematic and comprehensive evaluation of NAT methods against AT.

Neural machine translation suffers poor compositionality.

PhD in Computer Science, Zhejiang University and Westlake University (2020.9-2024.9).

Master of Science in Artificial Intelligence, University of Edinburgh (2017.9-2018.11).

Bachelor of Engineering in Electronic Information Engineering, Wuhan University (2013.9-2017.6).

Research Intern at Tencent AI lab (2022.10-2024.5).

Algorithem Engineer at Noah Ark's lab, Huawei (2018.12-2020.6).

Software Engineering Intern at VMware, Beijing (2016.9-2017.5).

Area Chair: ACL 2025

Conference Reviewer: ACL, EMNLP, COLING, ACL ARR, IJCAI.

Journal Reviewer: TMLR, JAIR, TACL, TASLP, TBD, TALLIP.

Outstanding Student Scholarship (Silver medal, Tencent Rhino-Bird Elite Program, 2024).

National Scholarship (Ministry of Education, 2023).

Dean's Medal (Westlake University, 2023).

Website's code is from Jon Barron.