avatar

Han Wu

Try to be an NLPer.

Bio

I am a researcher in Huawei Noah’s Ark Lab in Hong Kong. I finished my Ph.D degree at Department of Computer Science, City University of Hong Kong, supervised by Dr. Linqi Song . Prior to that, I obtained my bachelor degree from Xi’an Jiaotong University in 2018. My research interests are lied on larger language models for optimization problems and model merging/ensembling.

News

  • We have some intern openings based in HK, focusing on the research of LLMs. Drop me your CV if interesed.

  • Jul. 2024. We have two paper accepted in COLM, 2024.

  • Jun. 2024. We have one paper accepted in JSTSP, 2024.

  • Oct. 2023. The last paper produced in my PhD time was accepted in TASLP.

  • Oct. 2023. Two papers are accepted in EMNLP main conference, 2023.

  • Jul. 2023. Pass the oral defense!

  • May. 2023. We have one paper accepted in ACL findings, 2023.

  • Jan. 2023. We have on paper accepted in ICLR 2023.

  • Mar. 2022. We have one paper accepted by NAACL findings, 2022.

  • Feb. 2022. We have one paper accepted by ACL findings, 2022.

  • Sep. 2021. We have one paper accepted by EMNLP 2021.

Publications

(* indicates equal contribution.)

  • CoCA: Regaining Safety-awareness of Multimodal Large Language Models with Constitutional Calibration. [paper] [code]
    Jiahui Gao, Renjie Pi, Tianyang Han, Han Wu, Lanqing Hong, Lingpeng Kong, Xin Jiang and Zhenguo Li.
    The First Conference on Language Modeling, 2024. (COLM, 2024)

  • Learning From Correctness Without Prompting Makes LLM Efficient Reasoner. [paper] [code]
    Yuxuan Yao*, Han Wu*, Zhijiang Guo, Biyan Zhou, Jiahui Gao, Sichun Luo, Hanxu Hou, Xiaojin Fu and Linqi Song.
    The First Conference on Language Modeling, 2024. (COLM, 2024)

  • Towards Improving Interpretability of Language Model Generation Through a Structured Knowledge Discovery Approach. [paper] [code]
    Shuqi Liu, Han Wu, Guanzhi Deng, Jianshu Chen, Xiaoyang Wang and Linqi Song.
    IEEE Journal of Selected Topics in Signal Processing, 2024. (JSTSP, 2024)

  • Building Structure-Aware Conversational Semantic Role Labeling Model via Relational Graph Network. [paper] [code]
    Han Wu, Kun Xu and Linqi Song.
    IEEE Transactions on Audio, Speech and Language Processing, 2023. (TASLP, 2023)

  • Reconstruct Before Summarize: An Efficient Two-Step Framework for Condensing and Summarizing Meeting Transcripts. [paper] [code]
    Haochen Tan, Han Wu, Wei Shao, Xinyun Zhang, Mingjie Zhan, Zhaohui Hou, Ding Liang and Linqi Song.
    In Proceedings of Conference on Empirical Methods in Natural Language Processing, 2023. (EMNLP, 2023)

  • Fine-grained Conversational Decoding via Isotropic and Proximal Search. [paper] [code]
    Yuxuan Yao, Han Wu, Qiling Xu, and Linqi Song.
    In Proceedings of Conference on Empirical Methods in Natural Language Processing, 2023. (EMNLP, 2023)

  • VCSUM: A Versatile Chinese Meeting Summarization Dataset. [paper] [code]
    Han Wu, Mingjie Zhan, Haochen Tan, Zhaohui Hou, Ding Liang, Linqi Song.
    In Findings of the Annual Meeting of the Association for Computational Linguistics, 2023.(ACL Findings, 2023)

  • Learning Locality and Isotropy in Dialogue Modeling. [paper] [code]
    Han Wu, Haochen Tan, Mingjie Zhan, Gangming Zhao, Shaoqing Lu, Ding Liang and Linqi Song.
    In the Eleventh International Conference on Learning Representations, 2023.(ICLR 2023)

  • Zero-Shot Cross-Lingual Conversational Semantic Role Labeling. [paper] [code]
    Han Wu, Haochen Tan, Kun Xu, Shuqi Liu, Lianwei Wu and Linqi Song.
    In Findings of the North American Chapter of the Association for Computational Linguistics - Human Language Technologies (NAACL-HLT), 2022.(NAACL Findings, 2022)

  • A Sentence is Worth 128 Pseudo Tokens: A Semantic-Aware Contrastive Learning Framework for Sentence Embeddings. [paper]
    Haochen Tan, Wei Shao, Han Wu, Ke Yang and Linqi Song.
    In Findings of the Annual Meeting of the Association for Computational Linguistics, 2021.(ACL Findings, 2022)

  • CSAGN: Conversational Structure Aware Graph Network for Conversational Semantic Role Labeling. [paper] [code]
    Han Wu, Kun Xu and Linqi Song.
    In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2021. (EMNLP 2021)

  • Domain-Adaptive Pretraining Methods for Dialogue Understanding. [paper] [code]
    Han Wu, Kun Xu, Linfeng Song, Lifeng Jin, Haisong Zhang and Linqi Song.
    In Proceedings of the Annual Meeting of the Association for Computational Linguistics, 2021.(ACL 2021)

  • Conversational Semantic Role Labeling. [paper] [dataset]
    Kun Xu, Han Wu, Linfeng Song, Haisong Zhang, Linqi Song and Dong Yu.
    IEEE Transactions on Audio, Speech and Language Processing, 2021. (TASLP)

  • K-CSRL: Knowledge Enhanced Conversational Semantic Role Labeling. [paper]
    Boyu He, Han Wu, Congduan Li, Linqi Song and Weigang Chen.
    In Proceedings of 13th International Conference on Machine Learning and Computing, 2021. (ICMLC 2021)

  • Semantic Role Labeling Guided Multi-turn Dialogue Rewriter. [paper]
    Kun Xu, Haochen Tan, Linfeng Song, Han Wu, Haisong Zhang, Linqi Song and Dong Yu.
    In Proceedings of the Conference on Empirical Methods in Natural Language Processing, 2020. (EMNLP 2020)

Education

  • Ph.D. student, Department of Computer Science, City University of Hong Kong. Sep. 2019 - Aug. 2023.

  • B.Eng., School of Software Engineering, Xi’an Jiaotong University. Sep. 2014 - Jun. 2018.

Research Experience

  • Sensetime Research Institute., Research Intern with Mingjie Zhan and Ding Liang.
    Working on large-scale dialogue pre-training and dialogue summarization.
    Mar. 2022 - Feb. 2023, Beijing.

  • Tencent AI Lab, Research Intern with Dr. Kun Xu.
    Working on dialogue system and semantic parsing.
    Dec. 2019 - Oct. 2021, Shenzhen.

  • Dept. of Computer Science, CityU, Research Assistant with Dr. Linqi Song.
    Working on privacy-preserving recommender system.
    May. 2019 - Aug. 2019, Hong Kong.

  • Lab of Social Intelligence and Complex Data Processing, XJTU, Research Assistant with Prof. Yuan Rao.
    Working on event detection and storyline generation.
    Sep. 2018 - Apr. 2019, Xi’an.

Professional Service

Conference Reviewer: AAAI’(21-24), ACL’(22-24), NAACL’22, EMNLP’(21-24), IJCAI’23, NeurIPS’23-24, ICLR’24, ARR
Journal Reviewer: TASLP, TNNLS

Award

  • Outstanding Academic Performance Award(25%), CityU, 2021, 2022.

  • Research Tuition Scholarship (20%), CityU, 2021.

  • CityU PhD Studentship, 2019-2023.

  • First-class scholarship, XJTU, 2016, 2017.