[Google Scholar] [Github]



Bio:

Hi! I am Hai Ye. I am a Ph.D. candidate at the School of Computing, National University of Singapore (NUS), where I am working with prof. Hwee Tou Ng. My research is supported by AISG Ph.D. Fellowship. I obtained my bachelor degree from Beihang University, Beijing, China.

Research:

My research centers on Natural Language Processing and Machine Learning, with a current focus on Large Language Models:

  • Generalization and Robustness: I have explored methods to enhance the performance of language models across various domains and languages [1, 2]. My focus includes continual test-time adaptation [3] and iterative system updates driven by human feedback [4].
  • LLM Alignments: I am focusing on advancing LLMs through data-centric approaches. I am developing efficient methods for data synthesis to align models [5] and build robust judge and reward systems [6].

    • Contact: hye.me [at] outlook [dot] com

    Selected Papers

    (* indicates equal contribution)

     Preference-Guided Reflective Sampling for Aligning Language Models
     Hai Ye, Hwee Tou Ng
     In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024, Long paper.
     [paper] [code] [website]
     Multi-Source Test-Time Adaptation as Dueling Bandits for Extractive Question Answering
     Hai Ye, Qizhe Xie, Hwee Tou Ng
     In Proceedings of the Association for Computational Linguistic (ACL), 2023, Long paper.
     [paper] [code]
     Robust Question Answering against Distribution Shifts with Test-Time Adaptation: An Empirical Study
     Hai Ye, Yuyang Ding, Juntao Li, Hwee Tou Ng
     In Findings of the Association for Computational Linguistics: EMNLP 2022, Long paper.
     [paper] [code]
     On the Robustness of Question Rewriting Systems to Questions of Varying Hardness
     Hai Ye, Hwee Tou Ng, Wenjuan Han
     In Proceedings of the Association for Computational Linguistic (ACL), 2022, Long paper.
     [paper]
     On the Effectiveness of Adapter-based Tuning for Pretrained Language Model Adaptation
     Ruidan He*, Linlin Liu*, Hai Ye*, Qingyu Tan, Bosheng Ding, Liying Cheng, David Low, Lidong Bing, Luo Si
     In Proceedings of the Association for Computational Linguistic (ACL), 2021, Long paper.
     [paper]
     Feature Adaptation of Pre-Trained Language Models across Languages and Domains with Robust Self-Training
     Hai Ye, Qingyu Tan, Ruidan He, Juntao Li, Hwee Tou Ng, Lidong Bing
     In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2020, Long paper.
     [paper] [code]
     Unsupervised Domain Adaptation of a Pretrained Cross-Lingual Language Model
     Juntao Li, Ruidan He, Hai Ye, Hwee Tou Ng, Lidong Bing, Rui Yan
     In Proceedings of International Joint Conferences on Artificial Intelligence (IJCAI), 2020.
     [paper] [code]
     Jointly Learning Semantic Parser and Natural Language Generator via Dual Information Maximization
     Hai Ye, Wenjie Li, Lu Wang
     In Proceedings of the Association for Computational Linguistic (ACL), 2019, Long paper.
     [paper] [poster]
     Semi-Supervised Learning for Neural Keyphrase Generation
     Hai Ye, Lu Wang
     In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), 2018, Long paper.
     [paper]
     Interpretable Charge Predictions for Criminal Cases: Learning to Generate Court Views from Fact Descriptions
     Hai Ye*, Xin Jiang*, Zhunchen Luo*, Wenhan Chao
     In Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics (NAACL), 2018, Long paper, Oral.
     [paper] [Data] [Slides]
     Jointly Extracting Relations with Class Ties via Effective Deep Ranking
     Hai Ye, Wenhan Chao, Zhunchen Luo, Zhoujun Li
     In Proceedings of the Association for Computational Linguistic (ACL), 2017, Long paper.
     [paper] [code]

    Academic Services

  • Area Chair: ACL Rolling Review
  • Reviewer: ARR, ACL, NAACL, EMNLP, COLING, COLM 2024, NeurIPS 2024 and 2025, ICLR 2024 and 2025, ICML 2025
  • Volunteer: NAACL 2018

    • Last update: 2024