Percy Liang

Career

Percy Liang's CV records a postdoctoral position at Google in 2011–2012, after a PhD in computer science from the University of California at Berkeley. Liang joined the Stanford faculty as an Assistant Professor in 2012 and became an Associate Professor in 2019; the current homepage lists the title Professor of Computer Science, with a courtesy appointment in Statistics.⁷¹

Stanford CRFM's people page lists Liang as the center's director. Together AI's leadership page lists Liang among the company's founders, alongside Vipul Ved Prakash, Ce Zhang, Chris Ré and Tri Dao.²⁶

Research

Liang's homepage describes an interest in basic questions about learning and intelligence, and a commitment to efficient and reproducible research. Liang created CodaLab Worksheets, a platform that records the full provenance of an experiment from raw data to final results, and later created Marin, which practices "open development" by preregistering experiments and publishing them as they run.¹³

Liang is a co-author of the 2016 SQuAD paper, which introduced the Stanford Question Answering Dataset for reading comprehension, and of the 2021 report On the Opportunities and Risks of Foundation Models. Liang is first author of the 2022 HELM paper, which set out a taxonomy of scenarios and metrics for evaluating language models more transparently; the HELM framework is maintained as open source by Stanford CRFM.⁸⁹⁵¹⁰

Teaching

Liang's homepage lists courses including CS336 (Language Models from Scratch), CS324 (Advances in Foundation Models) and CS221 (Artificial Intelligence). The Spring 2026 CS336 site lists Liang and Tatsunori Hashimoto as the course's instructors.¹⁴

Work and contributions

• Marin Open-source code

  Created Marin, per Liang's homepage; the project's README names Stanford CRFM and Open Athena as its core collaborators.¹³

• HELM (Holistic Evaluation of Language Models) Open-source code

  First author of the HELM paper (2022); the open-source framework was created by Stanford CRFM.⁵¹⁰

• CS336: Language Modeling from Scratch Published research

  Listed as an instructor, with Tatsunori Hashimoto, of Stanford's CS336 course in Spring 2026.⁴

• SQuAD: Stanford Question Answering Dataset Open data

  Co-author, with Pranav Rajpurkar, Jian Zhang and Konstantin Lopyrev, of the 2016 paper that introduced the SQuAD reading-comprehension dataset.⁸

• CodaLab Worksheets Open-source code

  Created CodaLab Worksheets, a platform for running and managing experiments that keeps the full provenance from raw data to final results, per Liang's homepage.¹

Timeline

1. 2011Completed a PhD in computer science at the University of California at Berkeley.⁷
2. 2011Began a postdoctoral position at Google, held until 2012.⁷
3. 2012Joined Stanford University as an Assistant Professor.⁷
4. Jun 16, 2016Co-authored the SQuAD paper introducing the Stanford Question Answering Dataset.⁸
5. 2019Became an Associate Professor at Stanford University.⁷
6. Aug 16, 2021Co-authored the report On the Opportunities and Risks of Foundation Models.⁹
7. Nov 16, 2022Released the HELM paper on holistic evaluation of language models as first author.⁵
8. 2026Co-taught CS336, Language Modeling from Scratch, with Tatsunori Hashimoto in Spring 2026.⁴

Honors and recognition

• Microsoft Research Faculty Fellowship · Microsoft Research, 2014⁷
• Sloan Research Fellowship · Alfred P. Sloan Foundation, 2015⁷
• NSF CAREER Award · U.S. National Science Foundation, 2016⁷
• IJCAI Computers and Thought Award · International Joint Conferences on Artificial Intelligence, 2016⁷
• Best Paper Award (COLT) · Conference on Learning Theory, 2017⁷
• Best Paper Award (ICML) · International Conference on Machine Learning, 2017⁷
• Best Short Paper Award (ACL) · Association for Computational Linguistics, 2018⁷
• Presidential Early Career Award for Scientists and Engineers (PECASE) · U.S. Government, 2019⁷

Where they work

• Stanford University 40 people in the wiki
• Together AI 5 people in the wiki

Related people

• Tatsunori HashimotoAlso at Stanford University
• Dan JurafskyAlso at Stanford University
• Douwe KielaAlso at Stanford University
• Christopher ManningAlso at Stanford University
• Rohan TaoriAlso at Stanford University
• AI ExplainedTeaching and writing · Evaluation and benchmarks
• Emmanuel DupouxLanguage models · Evaluation and benchmarks
• Rowan CheungTeaching and writing · Language models
• Nanyun (Violet) PengLanguage models · Evaluation and benchmarks
• Eric Michael SmithEvaluation and benchmarks · Language models