教师名录
个人简介
I have been a faculty member at Shanghai Jiao Tong University (SJTU) since 2021. From 2016 to 2020, I worked as a researcher at the National Institute of Information and Communications Technology (NICT) in Japan, following an internship there in 2013. I am deeply grateful to my senseis at NICT—Masao Utiyama and Eiichiro Sumita—for their invaluable guidance, support, and inspiration throughout my time in Japan. From 2012 to 2016, I pursued a joint Ph.D. between SJTU and the French National Centre for Scientific Research (CNRS), under the supervision of Hai Zhao, Bao-Liang Lu, and Sabine Ploux. Prior to my doctoral studies, I studied and worked at the Xinjiang Branch of the Chinese Academy of Sciences from 2009 to 2012. I earned my bachelor's degree from Harbin Institute of Technology in 2009. I was born in Harbin, China, in 1985—the city I proudly call home.
Language Intelligence is a form of cognitive capability—shared by humans and increasingly emulated by machines—that enables agents to learn about the external world through natural language, construct highly abstract linguistic representations, and thereby understand, refine, and creatively transform that world. Our research seeks to explore and expand the frontiers of this intelligence across three interconnected domains:
★ Language Modeling: The mechanism and application. I started researching language models in 2013, which was a turning point when language models shifted from n-gram approaches to neural networks [EMNLP-2013].
•We have observed some interesting phenomena: [Overthinking], [Chain-of-Embedding], [Mixed-radix Rotary Position Embedding], etc.
•We have proposed [DeepMath-103K] with Tencent, [PolyMath] with Qwen Team, and [F2LLM] with Ant Group.
★ Computational Linguistics: An interdisciplinary field of computer science, linguistics, cognitive science, psychology, etc.
• We have provided the first evidence that [Neural Theory-of-Mind Networks] can spontaneously generalize high-order ToM, akin to human cognition.
• We have observed that LLM can solve concrete problems through [Meta-Reasoning].
★ Machine Translation (MT): I worked on MT for over ten years and will never give it up.
• We argue that the development trend of machine translation is gradually moving toward [Unsupervised Machine Translation].
• We are working on [Human-like Machine Translation] and [Multi-agent Machine Translation].
教授课程
CS3966: Natural Language Processing and Large Language Model (for the John Class), 2024-
CS3602: Natural Language Processing (for the CS and AI major), 2021-
CS438: Information Extraction, 2021-2023
CS247: Data Mining, 2021-2022
Advances and Challenges in Unsupervised Neural Machine Translation.
Rui Wang and Hai Zhao
16th conference of the European Chapter of the Association for Computational Linguistics (EACL-Tutorial), 2021
Syntax in End-to-End Natural Language Processing
Hai Zhao, Rui Wang, and Kehai Chen
The 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP-Tutorial), 2021
Domain Adaptation for Neural Machine Translation
Chenhui Chu and Rui Wang
The 16th China Conference on Machine Translation (CCMT-Tutorial), 2019
--You also refer to our survey paper in [COLING-2018]
论文发表
Recent Representative Works
[Neural Theory-of-Mind Networks]: Spontaneous High-Order Generalization in Neural Theory-of-Mind Networks. ArXiv. 2025
[DeepMath-103K]: DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning. ICLR 2026
[Overthinking]: Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs. ICML.2025
[Chain-of-Embedding]: Latent Space Chain-of-Embedding Enables Output-free LLM Self-Evaluation. ICLR.2025
[Human-like Machine Translation]: Exploring Human-Like Translation Strategy with Large Language Models. TACL.2024
资助项目
NSFC, 62176153, Multilingual Unsupervised Machine Translation, PI, 2022-2025
JSPS, 19K20354, Toward Unsupervised Machine Translation, PI, 2019-2020
We sincerely express our gratitude to Tencent, Ant Group, Alibaba, Hithink RoyalFlush, Baidu, and other companies for their generous financial support to our laboratory.
获奖信息
WMT-2024: 1st places in Translation into Low-Resource Languages of Spain[Paper]
WMT-2023: 1st places in Word-Level AutoCompletion[Paper]
WMT-2022: 1st places in Livonian<->English[Paper]
IWSLT-2022: 1st in the Simultaneous Speech Translation task [Results][Paper]
WMT-2020: 1st in three tasks (English->Chinese, Polish->English, and German-Upper Sorbian) [Results][Paper]
CoNLL-2019: 1st in the DM sub-task and the 2nd overall [Results][Paper]
WMT-2019: 1st in the only unsupervised MT task (German-Czech) [Results] [Paper]
WAT-2018: 1st places in Myanmar (Burmese) <- English [Results][Paper]
WMT-2018: 1st places in four tasks (English<->Estonian and English<->Finnish) [Results][Paper]
学术服务
Program Chair: MT Summit 2023 (Research Track)
Senior Area Chairs: NeurIPS (2026), ACL (2024, 2026), EACL (2026), EMNLP (2025-2026), NAACL (2025)
Area Chairs: ICML (2023-2026), NeurIPS (2022-2025), ICLR (2021-2026), EMNLP (2022), NAACL (2021), CoNLL (2021-2022), CCL (2018-2019)
Associate Editors: ACL Rolling Review (2024), IEICE Transactions on Information and Systems (2022-2024)
Standing Reviewers: CL (2020-2024) and TACL (2020-2024)
English
