Zonghan Yang (杨宗瀚) Final-year doctoral student Department of Computer Science and Technology Institute for AI Industry Research Tsinghua University Email: yangzh20@mails.tsinghua.edu.cn [Featured Publications] [Service] [CV (2024.10)] [Twitter]

Welcome! I am a final-year Computer Science PhD student at Tsinghua University. I am proudly advised by Prof. Yang Liu and affiliated with both THUNLP and AIR. I also obtained my B.Eng in 2020 from Tsinghua University, majoring in Computer Science and Technology.

I am studying agents that automate tasks ranging from daily activities to creative endeavors. Previously, I've also designed algorithms to improve the adaptivity and responsibility of LLMs with control theories and test-time compute (a Twitter thread here). I am constantly driven by the mysteries of the unknown, and commited to conducting goal-driven research that creates a positive impact.

I am actively seeking postdoc and industrial research opportunities! Don't hesitate to reach out to me if you are interested in my work or want to get in touch with me.

• ReAct Meets ActRe: Autonomous Annotation of Agent Trajectories for Contrastive Self-Training
  Zonghan Yang, Peng Li, Ming Yan, Ji Zhang, Fei Huang, and Yang Liu.
  COLM 2024.

      @inproceedings{yang2024react,
        title={ReAct Meets ActRe: Autonomous Annotation of Agent Trajectories for Contrastive Self-Training},
        author={Yang, Zonghan and Li, Peng and Yan, Ming and Zhang, Ji and Huang, Fei and Liu, Yang.},
        booktitle={CoLM},
        year={2024},
      }
    

[bib] [arXiv]



• Towards Unified Alignment Between Agents, Humans, and Environment
  Zonghan Yang*, An Liu*, Zijun Liu*, Kaiming Liu, Fangzhou Xiong, Yile Wang, Zeyuan Yang, Qingyuan Hu, Xinrui Chen, Zhenhe Zhang, Fuwen Luo, Zhicheng Guo, Peng Li, Yang Liu.
  ICML 2024; LLMAgents Workshop @ ICLR 2024.

      @article{yang2024towards,
        title = {Towards Unified Alignment Between Agents, Humans, and Environment},
        author = {Yang, Zonghan and Liu, An and Liu, Zijun and Liu, Kaiming and Xiong, Fangzhou and Wang, Yile and Yang, Zeyuan and Hu, Qingyuan and Chen, Xinrui and Zhang, Zhenhe and Luo, Fuwen and Guo, Zhicheng and Li, Peng and Liu, Yang},
        journal={arXiv preprint arXiv:2402.07744},
        year = {2024}
      }
    

[bib] [arXiv] [project webpage] [Twitter thread]



• Scaffolding Coordinates to Promote Vision-Language Coordination in Large Multi-Modal Models
  Xuanyu Lei, Zonghan Yang, Xinrui Chen, Peng Li, Yang Liu.
  Wordplay Workshop @ ACL 2024.

      @article{lei2024scaffolding,
        title={Scaffolding Coordinates to Promote Vision-Language Coordination in Large Multi-Modal Models},
        author={Lei, Xuanyu and Yang, Zonghan and Chen, Xinrui and Li, Peng and Liu, Yang},
        journal={arXiv preprint arXiv:2402.12058},
        year={2024},
      }
    

[bib] [arXiv] [project webpage] [code] [Twitter thread]



• Unified Detoxifying and Debiasing in Language Generation via Inference-time Adaptive Optimization
  Zonghan Yang, Xiaoyuan Yi, Peng Li, Yang Liu, Xing Xie.
  ICLR 2023.

        @inproceedings{
            yang2023unified,
            title={Unified Detoxifying and Debiasing in Language Generation via Inference-time Adaptive Optimization},
            author={Yang, Zonghan and Yi, Xiaoyuan and Li, Peng and Liu, Yang and Xie, Xing},
            booktitle={The Eleventh International Conference on Learning Representations },
            year={2023},
            url={https://openreview.net/forum?id=FvevdI0aA_h}
          }
      

[bib] [arXiv] [code]



• On Robust Prefix-Tuning for Text Classification
  Zonghan Yang, Yang Liu.
  ICLR 2022.

      @inproceedings{
        yang2022on,
        title={On Robust Prefix-Tuning for Text Classification},
        author={Zonghan Yang and Yang Liu},
        booktitle={International Conference on Learning Representations},
        year={2022},
        url={https://openreview.net/forum?id=eBCmOocUejf}
      }
    

[bib] [arXiv] [code] [slides] [blog]



• Reducing Word Omission Errors in Neural Machine Translation: A Contrastive Learning Approach
  Zonghan Yang, Cheng Yong, Yang Liu, Maosong Sun.
  ACL 2019 (short paper).

      @inproceedings{yang-etal-2019-reducing,
        title = "Reducing Word Omission Errors in Neural Machine Translation: A Contrastive Learning Approach",
        author = "Yang, Zonghan  and
          Cheng, Yong  and
          Liu, Yang  and
          Sun, Maosong",
        booktitle = "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
        month = jul,
        year = "2019",
        address = "Florence, Italy",
        publisher = "Association for Computational Linguistics",
        url = "https://aclanthology.org/P19-1623",
        doi = "10.18653/v1/P19-1623",
        pages = "6191--6196",
      }
    

[bib] [Implementation in OpenNMT-tf]

• Reviewer (primary & secondary) for ICLR, ICML, NeurIPS, ACL, EMNLP, EACL, TACL, AAAI, IJCAI
• Co-chair for Student Research Workshop (SRW) in AACL-IJCNLP 2022: [Website] [Proceedings]
• Editor for THUMT Research Blog
• Maintainer for Machine Translation Reading List (inactive now...)
• Teaching Assistant:
  1. (40240432) Formal Language & Automaton: 2021-2022 Fall, 2022-2023 Fall, 2023-2024 Fall
  2. (14204002) Progressive English Reading & Writing: 2020-2021 Fall, 2020-2021 Spring