LAVINE
≡
profile photo

Wen Lai (赖文)

I am a final-year PhD student in School of Computation, Information and Technology at Technical University of Munich (TUM) under the supervision of Prof. Alexander Fraser and mentored by Prof. Ivan Titov. I am affiliated with and supported by the Munich Center for Machine Learning. I initially start my PhD in Center for Information and Language Processing (CIS) at Ludwig-Maximilians-Universität München (LMU Munich) from April 2021 to March 2024 and later transferred to TUM along with my supervisor.

emailEmail googleGoogle Scholar githubGithub tweetTweet tweetLinkedin tweetBlog

Research Interests

  • Machine Translation: Low-resource MT, Domain Adaptation, Multilingual MT, LLM for MT
  • Machine Learning: Meta-Learning, Transfer Learning, Contrastive Learning
  • Multimodal: Large Vision-Language Models

    • News

  • 02/2024, one paper in ArXiv: DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detection.

  • 02/2024, one paper in ArXiv: Joint Localization and Activation Editing for Low-Resource Fine-Tuning.

  • 10/2024, I am visiting the NLP Group at University of Edingburg from Octorber 2024.

  • 09/2024, One paper accepted to EMNLP 2024.

  • 05/2024, One paper accepted to ACL 2024.

  • 04/2024, I will be working at Technical University of Munich (TUM) as a PhD Student from April 2024.

  • 11/2023, One paper in Arxiv titled Extending Multilingual Machine Translation through Imitation Learning.

  • 10/2023, One paper accepted to EMNLP2023.

  • 07/2023, I will be working at Bosch Center for Artificial Intelligence as a research intern from September 2023.

  • 05/2023, One paper in Arxiv titled "Mitigating Data Imbalance and Representation Degeneration in Multilingual Machine Translation ".

  • 03/2023, I gave a presentation at the Helsinki NLP Group called "Domain Adaptation for Machine Translation in Few-Shot Setting". Check out the video and the slides.

  • 02/2023, Invited to serve as reviewer for LoResMT 2023.

  • 11/2022, Invited to serve as reviewer for EACL 2023.

  • 10/2022, One paper accepted to EMNLP2022.

  • 08/2022, One paper accepted to COLING2022.

  • 06/2022, Invited to serve as reviewer/PC member for LoResMT2022 and WMT2022.

  • 12/2021, One paper in Arxiv.

  • 10/2021, One paper in WMT2021.

  • 04/2021, Join CIS Group at LMU Munich under the supervision of Prof. Alexander Fraser.

  • 10/2020, Start to remote visit in Alex's Group.

  • 05/2020, Defended my Master Thesis "Neural Machine Translation for Low-Resource Languages".

    • Publications

  • DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detection
          Yingli Shen*, Wen Lai*, Shuo Wang, Xueren Zhang, Kangyang Luo, Alexander Fraser and Maosong Sun
          arxiv.2502.11546 | paper | code | dataset (* Equal contribution)

  • Joint Localization and Activation Editing for Low-Resource Fine-Tuning
          Wen Lai, Alexander Fraser, Ivan Titov
          arxiv.2502.01179 | paper | code

  • Style-Specific Neurons for Steering LLMs in Text Style Transfer
          Wen Lai, Viktor Hangya, Alexander Fraser
          EMNLP 2024 | paper | code | poster | slides

  • LLMs Beyond English: Scaling the Multilingual Capability of LLMs with Cross-Lingual Feedback
          Wen Lai, Mohsen Mesgar, Alexander Fraser
          ACL 2024 | paper | code | poster | slides

  • Extending Multilingual Machine Translation through Imitation Learning
          Wen Lai, Viktor Hangya, Yingli Shen, Alexander Fraser
          ArXiv:2311.08538 | paper

  • Mitigating Data Imbalance and Representation Degeneration in Multilingual Machine Translation
          Wen Lai, Alexandra Chronopoulou, Alexander Fraser
          EMNLP 2023 | paper | code | poster | slides

  • m4Adapter: Multilingual Multi-Domain Adaptation for Machine Translation with a Meta-Adapter
          Wen Lai, Alexandra Chronopoulou, Alexander Fraser
          EMNLP 2022 | paper | code | blog | poster | slides

  • Improving Both Domain Robustness and Domain Adaptability in Machine Translation
          Wen Lai, Jindřich Libovický, Alexander Fraser
          COLING 2022 | paper | code | blog | poster | slides

  • The LMU Munich System for the WMT 2021 Large-Scale Multilingual Machine Translation Shared Task
          Wen Lai, Jindřich Libovický, Alexander Fraser
          WMT 2021 | paper | poster

    • Honors and Awards

  • 06/2021, Facebook Compute Grant, Large-Scale Multilingual Machine Translation Shared Tasks

  • 06/2020, Outstanding Master Thesis paper at Minzu University of China

  • 08/2018, CWMT2018: Uyghur-Chinese Track (2nd), Tibetan-Chinese Track (3rd)

    • Selected Students

  • Marko Bierlich: Analyzing the Anisotropy Phenomenon in Multilingual Machine Translation Models. (LMU Munich, Bachelor Thesis, 09/2022-12/2022)

  • Bartłomiej Jezierski: Transliteration-Aided Machine Translation for Related Low-Resource Languages: Upper and Lower Sorbian. (Technical University of Munich, Guided Research, 04/2022-12/2022)

    • Services

  • ACL Rolling Reviewer
  • NAACL 2025
  • EMNLP 2023
  • EACL 2023
  • WMT 2023, 2022
  • LoresMT 2023, 2022
    • Pageviews
    Welcome to Wen Lai's website
    No. Visitor Since Jan 2022. Powered by w3.css