Hai Hu

Hai Hu is an Assistant Professor in the Translation Department at Shanghai Jiao Tong University (SJTU), specializing in Computational Linguistics and Natural Language Processing (NLP). He leads the Computational Linguistics (CL) lab at SJTU's School of Foreign Languages, focusing on computational modeling of human language aspects (syntax, semantics, pragmatics), NLP evaluation with linguistically oriented benchmarks, and comparisons between human language acquisition and machine learning.

Education: PhD in (Computational) Linguistics from Indiana University Bloomington (IU), advised by Larry Moss and Sandra Kübler. BA in English and MA in English Linguistics from Renmin University of China. Exchange at Universität Tübingen.

Research interests include natural language inference, Chinese linguistics, language variation, corpus annotation, and stylistic properties of translations. The CL lab actively recruits students with backgrounds in linguistics, computer science, translation, and cognitive science. Lab members work on projects involving deep learning servers for training state-of-the-art language models.

Key contributions include the CoLAC corpus (Chinese linguistic acceptability), ZhoBLiMP (systematic Chinese NLP benchmarks), and MELA (multilingual acceptability evaluation). His work explores cross-lingual transfer learning in models like XLM-R and evaluates language models through linguistically rigorous tests.

Awards: Highlight Paper Award at CCL 2024 for work on conversational implicature in Chinese sitcoms. Extensive publication record in top venues like Computational Linguistics, ACL, and COLING.