Jian Wang (王健) Postdoctoral Fellow Department of Computing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China Email: jian51.wang (at) polyu.edu.hk   or                jwanglvy (at) gmail.com Google Scholar   |   Github   |   Twitter

Hi there! I am a Postdoctoral Fellow at PolyU NLP Group. I earned my PhD degree from The Hong Kong Polytechnic University in 2024, supervised by Prof. Maggie Wenjie Li. During the spring and summer of 2024, I visited the SLED Lab at University of Michigan, where I was so honored to work with Prof. Joyce Chai.

Prior to my PhD, I earned MSc and BEng degrees from the School of Computer Science and Engineering, South China University of Technology. During my graduate studies, I was fortunate to be co-advised by Prof. Min Yang at SIAT, Chinese Academy of Sciences.

🔥 Our group has 2 ~ 3 PhD positions for Fall 2026 who are passionate about LLM agents, embodied reasoning, situated interactions, and related areas. If you are interested, feel free to email me your CV and a brief self-intro!

News

[Aug. 2025] 🔥 Got one paper accepted by EMNLP 2025.

[July 2025] 🔥 Got one paper accepted by ACM MM 2025.

[Jun. 2025] 🔥 Our ACL paper was selected as an oral presentation (Top 8%)!

[May. 2025]   Got three papers accepted by ACL 2025! Welcome to have a chat in Vienna.

[Feb. 2025]   Thrilled to be awarded the PolyU Postdoc Matching Fund.

[Oct. 2024]   Will be working as a Postdoc at PolyU NLP Group.

[Sep. 2024]   Defended my PhD thesis. Huge thanks to my supervisor and committee members!

[May 2024]   Got one paper accepted by ACL 2024! Welcome to have a chat in Bangkok.

[Dec. 2023]   Will be visiting the SLED Lab at University of Michigan, starting from Feb. 2024. Very excited!

[Oct. 2023]   Got two papers accepted by EMNLP 2023! Will be heading to Singpore in December.

[Jul. 2023]   Will be heading to Toronto for ACL 2023. The first time to attend an in-person academic conference during my PhD and to visit abroad during my life!

Research

My research interests are in natural language processing (NLP), large language models (LLMs), and AI more broadly. Recently, I have been focusing on the following directions:

• Reasoning and Reliability of LLMs: test-time scaling (arXiv'25a, arXiv'25b), safety alignment (ACL'25), latent reasoning (arXiv'25c), etc.
• Digital and Embodied Agents: embodied reasoning, environment adaptation, agentic RL (arXiv'25), tutoring agents (ACL-Findings'25a), task planning(ACL-Findings'25b), etc.
• Conversational AI: proactive dialogue systems (TNNLS'23, TOIS'24, ACL-Findings'23), dialogue consistentcy (ACL'24), personalized role-playing (EMNLP'23), emotional support conversation (AAAI'24), etc.

👉 Welcome highly motivated students to collaborate on research projects related to (but not limited to) the topics mentioned above. If you are interested, please feel free to drop me an email.

Selected Publications

(see full list in Google Scholar) (*: Equal contribution; ^†: Corresponding author)

Preprints

• SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution

  Hanlin Wang, Chak Tou Leong, Jiashuo Wang, Jian Wang^†, and Wenjie Li
  arXiv:2505.20732, 2025
  arXiv | Github

• Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning

  Xinghao Chen, Anhao Zhao, Heming Xia, Xuan Lu, Hanlin Wang, Yanjun Chen, Wei Zhang, Jian Wang^†, Wenjie Li, and Xiaoyu Shen^†
  arXiv:2505.16782, 2025
  arXiv | Github

• Probabilistic Optimality for Inference-time Scaling

  Youkang Wang*, Jian Wang*, Rubing Chen, Xiao-Yong Wei, and Qing Li.
  arXiv:2506.22376, 2025
  arXiv | Github

• Scaling over Scaling: Exploring Test-Time Scaling Plateau in Large Reasoning Models

  Jian Wang, Boyan Zhu, Chak Tou Leong, Yongqi Li^†, and Wenjie Li.
  arXiv:2505.20522, 2025
  arXiv

2025

• Benchmarking for Domain-Specific LLMs: A Case Study on Academia and Beyond

  Rubing Chen, Jiaxin Wu, Jian Wang, Xulu Zhang, Wenqi Fan, Chenghua Lin, Xiao-Yong Wei, Qing Li.
  Findings of the Association for Computational Linguistics: EMNLP (EMNLP-Findings), 2025
  arXiv | Github

• Towards Harmless Multimodal Assistants with Blind Preference Optimization

  Yongqi Li, Lu Yang, Jian Wang, Runyang You, Wenjie Li, and Liqiang Nie.
  The 33rd ACM International Conference on Multimedia (ACM MM), 2025.
  arXiv | Github

• Training Turn-by-Turn Verifiers for Dialogue Tutoring Agents: The Curious Case of LLMs as Your Coding Tutors

  Jian Wang, Yinpei Dai, Yichi Zhang, Ziqiao Ma, Wenjie Li, and Joyce Chai.
  Findings of the Association for Computational Linguistics: ACL (ACL-Findings), 2025
  arXiv | Paper | Github

• STeCa: Step-level Trajectory Calibration for LLM Agent Learning

  Hanlin Wang, Jian Wang^†, Chak Tou Leong, and Wenjie Li.
  Findings of the Association for Computational Linguistics: ACL (ACL-Findings), 2025
  arXiv | Paper | Github

• Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region

  Chak Tou Leong, Qingyu Yin, Jian Wang^†, and Wenjie Li.
  The 63rd Annual Meeting of the Association for Computational Linguistics (ACL), 2025.
  arXiv | Paper | Github

2021 - 2024 (PhD period)

• Instruct Once, Chat Consistently in Multiple Rounds: An Efficient Tuning Framework for Dialogue

  Jian Wang, Chak Tou Leong, Jiashuo Wang, Dongding Lin, Wenjie Li, and Xiao-Yong Wei.
  The 62nd Annual Meeting of the Association for Computational Linguistics (ACL), 2024.
  arXiv | Paper | Github

• Target-constrained Bidirectional Planning for Generation of Target-oriented Proactive Dialogue

  Jian Wang, Dongding Lin, and Wenjie Li.
  ACM Transactions on Information Systems (TOIS), 2024.
  arXiv | Paper | Github

• SCREEN: A Benchmark for Situated Conversational Recommendation

  Dongding Lin*, Jian Wang*, Chak Tou Leong, and Wenjie Li.
  The 32nd ACM International Conference on Multimedia (ACM MM), 2024.
  Paper | Github

• COOPER: Coordinating Specialized Agents towards a Complex Dialogue Goal

  Yi Cheng, Wenge Liu, Jian Wang, Chak Tou Leong, Yi Ouyang, Wenjie Li, Xian Wu, and Yefeng Zheng.
  AAAI Conference on Artificial Intelligence (AAAI), 2024.
  arXiv | Paper | Github

• Target-oriented Proactive Dialogue Systems with Personalization: Problem Formulation and Dataset Curation

  Jian Wang, Yi Cheng, Dongding Lin, Chak Tou Leong, and Wenjie Li.
  The 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023.
  arXiv | Paper | Github

• Self-Detoxifying Language Models via Toxification Reversal

  Chak Tou Leong, Yi Cheng, Jiashuo Wang, Jian Wang, and Wenjie Li.
  The 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2023.
  arXiv | Paper | Github

• Dialogue Planning via Brownian Bridge Stochastic Process for Goal-directed Proactive Dialogue

  Jian Wang*, Dongding Lin*, and Wenjie Li.
  Findings of the Association for Computational Linguistics: ACL (ACL-Findings), 2023.
  arXiv | Paper | Github

• A Target-driven Planning Approach for Goal-directed Dialog Systems

  Jian Wang, Dongding Lin, and Wenjie Li.
  IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2023.
  Paper | PDF | Github

2018 - 2021 (Before PhD)

• Template-guided Clarifying Question Generation for Web Search Clarification

  Jian Wang and Wenjie Li.
  ACM International Conference on Information and Knowledge Management (CIKM), 2021.
  Paper | Slides | Github

• Dual Dynamic Memory Network for End-to-End Multi-turn Task-oriented Dialog Systems

  Jian Wang, Junhao Liu, Wei Bi, Xiaojiang Liu, Kejing He, Ruifeng Xu, and Min Yang.
  International Conference on Computational Linguistics (COLING), 2020.
  Paper | Github

• Improving Knowledge-aware Dialogue Generation via Knowledge Base Question Answering

  Jian Wang, Junhao Liu, Wei Bi, Xiaojiang Liu, Kejing He, Ruifeng Xu, and Min Yang.
  AAAI Conference on Artificial Intelligence (AAAI), 2020.
  arXiv | Paper | Github

• Topic Discovery by Spectral Decomposition and Clustering with Coordinated Global and Local Contexts

  Jian Wang, Kejing He, and Min Yang.
  International Journal of Machine Learning and Cybernetics (JMLC), 2020.
  Paper | PDF

Experiences

• 2024/02 - 2024/06: Visiting PhD student at University of Michigan (Advisor: Prof. Joyce Chai)
• 2020/11 - 2021/08: Research assistant at PolyU (Advisor: Prof. Maggie Wenjie Li)
• 2020/07 - 2020/10: Research intern at Rulai Inc. (Advisor: Dr. Yi Zhang)
• 2018/08 - 2019/09: Visiting graduate student at SIAT, Chinese Academy of Sciences (Advisor: Prof. Min Yang)

Honors and Awards

• Postdoc Matching Fund (Awarded by PolyU)
• Research Postgraduate Student Innovation Award (Awarded by HKU)
• Third Prize of 2021 Language and Intelligence Challenge (Awarded by CCF and CIPS)
• AAAI 2020 Student Scholarship
• National Scholarship of China

Professional Activities

Academic Services

• Area Chair: ACL Rolling Review (ARR), AACL (2025)
• Program Committee Member: NeurIPS, ACL, EMNLP, AAAI, COLING
• Journal Reviewer: IEEE Transactions on Neural Networks and Learning Systems (TNNLS), IEEE Transactions on Knowledge and Data Engineering (TKDE), IEEE Transactions on Multimedia (TMM), ACM Transactions on Information Systems (TOIS)

Teaching Assistant

• COMP 6709: Advanced Natural Language Processing, Spring 2023, PolyU
• COMP 6709: Advanced Natural Language Processing, Spring 2022, PolyU
• COMP 2021: Object-Oriented Programming, Fall 2022, PolyU
• COMP 2021: Object-Oriented Programming, Fall 2021, PolyU
• Algorithm Design and Analysis, Spring 2018, SCUT
• Advanced Programming Language (C++), Fall 2017, SCUT

Open-Source Projects

• 📖 Paper-Reading-ConvAI : Paper reading list in conversational AI.
• 🐳 PyLoader: An asynchronous Python dataloader for loading big datasets.

Useful Resources

• 📝 Highly Opinionated Advice on How to Write ML Papers (by Neel Nanda).
• 📝 A long talk about "Writing Perfect Papers" (Slides | Youtube | Bilibili) (by Prof. Baochun Li).
• 🍷 Lilian Weng's Blog: Lilian Weng's blog with broad learning notes about DL/NLP.
• 📰 Sebastian Ruder's Blog: Sebastian Ruder's blog with deep insights about NLP/ML/DL.
• 📸 Jay Alammar's Blog: Visualizations and animations for many language models (Transformer, BERT, etc).
• 📡 Scientific Spaces: Jianlin Su's blog (in Chinese) about ML|NLP theories with deep explorations.
• 👩‍🎓 Global Equality for PhDs: Resources to help global equality for PhDs in NLP/AI, contributed by Zhijing Jin.

Misc.

• I enjoy many ball games such as basketball, table tennis, billiards, and badminton. I also spend my leisure time on variety shows, literature books, and movies.

• I dream of traveling around China and many places of interest around the world someday.