- Senior Staff Research Scientist, Google1
- Associate Professor (on leave), Computer Science Department; director of PLUSLAB, University of California, Los Angeles1
Nanyun (Violet) Peng is a Senior Staff Research Scientist at Google and an Associate Professor (on leave) in the Computer Science Department of the University of California, Los Angeles, where Peng directs PLUSLAB (Peng's Language Understanding and Synthesis Lab). The lab's research covers controllable and creative language generation, multimodal foundation models, automatic evaluation of foundation models, and low-resource, multilingual natural language understanding. Before UCLA, Peng was a Research Assistant Professor in computer science at the University of Southern California and a Research Lead at its Information Sciences Institute. Peng is a co-author of Plan-And-Write: Towards Better Automatic Storytelling, which first plans a storyline and then generates a story from it.12
At a glance
- Current roles
- Senior Staff Research Scientist +1 more
- Sector
- Company
- Education
- Johns Hopkins, UCLA, USC
- Work listed
- 1
- Honors
- 3
- Sources cited
- 4
- Profile reviewed
Career
Peng completed a PhD in computer science at Johns Hopkins University's Center for Language and Speech Processing, then spent three years at the University of Southern California as a research assistant professor of computer science and a research lead at its Information Sciences Institute. Peng then joined UCLA's computer science department, where Peng directs PLUSLAB and received tenure in 2024, and is now a Senior Staff Research Scientist at Google while on leave from UCLA.1
In professional service, Peng was elected to the NAACL board in December 2024 and was named a program chair for ICLR 2025 and for EMNLP 2025, after earlier roles such as senior area chair for AACL 2022.1
Research
Peng's group aims to make NLP systems better at communicating with people, with a particular focus on creativity in AI. Its work spans controllable and creative language generation, multimodal foundation models, automatic evaluation of foundation models, and low-resource and multilingual language understanding. An early example is Plan-And-Write (2018), which generates stories by first planning a storyline from a title.12
Peng's NSF CAREER project, which began in 2024, studies insertion-based language models that generate text by repeatedly inserting words into an incomplete context rather than writing strictly left to right. Papers reported under the award include work on whether language models can write human-level narratives, on controlling language models, and on melody-constrained lyrics editing.341
Work and contributions
Plan-And-Write: Towards Better Automatic Storytelling Published research
Co-author of the 2018 paper proposing a hierarchical plan-and-write framework for open-domain story generation from a title, comparing dynamic and static storyline planning.2
Timeline
- Co-authored "Plan-And-Write: Towards Better Automatic Storytelling".2
- Received a Google Research Scholar award.1
- Named a program chair for ICLR 2025 and EMNLP 2025.1
- Received an NSF CAREER award for research on insertion-based natural language generation.31
- Received tenure at UCLA.1
- PLUSLab papers received three Outstanding Paper Awards at EMNLP 2024.1
- Elected as a board member of NAACL.1
Other documented roles
Roles the sources document that are not confirmed as current.
- Research Assistant Professor, Computer Science; Research Lead, Information Sciences Institute, University of Southern California1
Honors and recognition
Where they work
Related people
- Diogo AlmeidaAlso at Google
- Michael CollinsAlso at Google
- Margaret MitchellAlso at Google
- Nicholas CarliniAlso at Google
- Sergey BrinAlso at Google
- Douwe KielaLanguage models · Evaluation and benchmarks
- Mark ChenLanguage models · Multimodal models
- Ethan PerezEvaluation and benchmarks · Language models
- Ilya SutskeverLanguage models · Multimodal models
- Thomas ScialomLanguage models · Evaluation and benchmarks
Suggest a correction or removal
Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.
References
4 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026