Wanli Yang
Wanli Yang

Wanli Yang

PhD Student
Institute of Computing Technology
Chinese Academy of Sciences
Beijing, China

About Me

Hi! I'm a third-year PhD student at Institute of Computing Technology, Chinese Academy of Sciences (ICT, CAS), fortunate to be advised by Prof. Fei Sun and Prof. Xinran Liu. Previously, I obtained my B.E. in Software Engineering from Nankai University.

I study how AI systems can improve reliably: safely updating parametric knowledge through model editing, eliciting latent knowledge through reinforcement learning, and more broadly investigating whether and how autonomous agents can improve the AI R&D process itself.

News
2026

Together with the Meituan LongCat Team, we released Beyond Final Scores, a technical report evaluating agents for autonomous AI R&D.

Aug

Our new preprint explores how reinforcement learning unlocks parametric knowledge in LLMs.

May

One paper was accepted by ICLR 2026.

Jan
2025

One paper was accepted by ACL 2025.

May
2024

One paper was accepted by EMNLP 2024.

Sep

Two papers were accepted by ACL 2024.

May
Preprints View all
Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development
Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development

Yiwei Li*, Wanli Yang*, Hexiang Tan*, Xiangzhou Huang, Zhengyu Chen, Ziran Li, Borun Chen, Shanglin Lei, Huaisheng Zhu, Hao Tian, Fei Sun, Xunliang Cai, Jingang Wang (* equal contribution)

Technical report, 2026

[Paper] [Website]

Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development
Beyond Final Scores: A Systematic Evaluation of Agents for Long-Horizon AI Research and Development

Yiwei Li*, Wanli Yang*, Hexiang Tan*, Xiangzhou Huang, Zhengyu Chen, Ziran Li, Borun Chen, Shanglin Lei, Huaisheng Zhu, Hao Tian, Fei Sun, Xunliang Cai, Jingang Wang (* equal contribution)

Technical report, 2026

[Paper] [Website]

Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs
Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs

Wanli Yang, Hongyu Zang, Junwei Zhang, Wenjie Shi, Du Su, Jingang Wang, Xueqi Cheng, Fei Sun

Preprint

[Paper]

Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs
Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs

Wanli Yang, Hongyu Zang, Junwei Zhang, Wenjie Shi, Du Su, Jingang Wang, Xueqi Cheng, Fei Sun

Preprint

[Paper]

Selected Publications View all
Fine-tuning Done Right in Model Editing
Fine-tuning Done Right in Model Editing

Wanli Yang, Rui Tang, Hongyu Zang, Du Su, Qi Cao, Jingang Wang, Huawei Shen, Xueqi Cheng, Fei Sun

ICLR 2026

[Paper] [Code]

Fine-tuning Done Right in Model Editing
Fine-tuning Done Right in Model Editing

Wanli Yang, Rui Tang, Hongyu Zang, Du Su, Qi Cao, Jingang Wang, Huawei Shen, Xueqi Cheng, Fei Sun

ICLR 2026

[Paper] [Code]

The Mirage of Model Editing: Revisiting Evaluation in the Wild
The Mirage of Model Editing: Revisiting Evaluation in the Wild

Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Qi Cao, Dawei Yin, Huawei Shen, Xueqi Cheng

ACL 2025 Main & KnowFM@ACL 2025 Workshop Best Paper 🏅

[Paper] [Code] [Website]

The Mirage of Model Editing: Revisiting Evaluation in the Wild
The Mirage of Model Editing: Revisiting Evaluation in the Wild

Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Qi Cao, Dawei Yin, Huawei Shen, Xueqi Cheng

ACL 2025 Main & KnowFM@ACL 2025 Workshop Best Paper 🏅

[Paper] [Code] [Website]

The Fall of ROME: Understanding the Collapse of LLMs in Model Editing
The Fall of ROME: Understanding the Collapse of LLMs in Model Editing

Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Du Su, Dawei Yin, Huawei Shen

EMNLP 2024 Findings

[Paper] [Code] [Website]

The Fall of ROME: Understanding the Collapse of LLMs in Model Editing
The Fall of ROME: Understanding the Collapse of LLMs in Model Editing

Wanli Yang, Fei Sun, Jiajun Tan, Xinyu Ma, Du Su, Dawei Yin, Huawei Shen

EMNLP 2024 Findings

[Paper] [Code] [Website]

🦋🌪️ The Butterfly Effect of Model Editing: Few Edits Can Trigger Large Language Models Collapse
🦋🌪️ The Butterfly Effect of Model Editing: Few Edits Can Trigger Large Language Models Collapse

Wanli Yang, Fei Sun, Xinyu Ma, Xun Liu, Dawei Yin, Xueqi Cheng

ACL 2024 Findings

[Paper] [Code] [Website]

🦋🌪️ The Butterfly Effect of Model Editing: Few Edits Can Trigger Large Language Models Collapse
🦋🌪️ The Butterfly Effect of Model Editing: Few Edits Can Trigger Large Language Models Collapse

Wanli Yang, Fei Sun, Xinyu Ma, Xun Liu, Dawei Yin, Xueqi Cheng

ACL 2024 Findings

[Paper] [Code] [Website]

Internships & Education

Internships

Education

  • Institute of Computing Technology, Chinese Academy of Sciences

    PhD Student 2024 – present

  • Nankai University

    B.E. in Software Engineering 2020 – 2024

Academic Service

Reviewing

Reviewer for ACL Rolling Review (ARR; including ACL and EMNLP); ICLR; NeurIPS.

Recognized in ARR Great Reviewers (Jan. 2026).

Tutorials

Model Editing in LLMs: Advances and Challenges

NLPCC 2025 · Xinjiang, China Aug. 2025
Invited tutorial presenter. With Dr. Junfeng Fang.