Noah A. Smith

Career

Smith earned a PhD in computer science at Johns Hopkins University, advised by Jason Eisner. Smith was previously an associate professor of language technologies and machine learning in Carnegie Mellon University's School of Computer Science, before joining the University of Washington's Paul G. Allen School.¹

At UW, Smith is the inaugural Vice Provost for Artificial Intelligence and holds the Charles and Lisa Simonyi Endowed Chair for Artificial Intelligence and Emerging Technologies. Smith is also Senior Director of NLP Research at the Allen Institute for AI and principal investigator of the NSF- and NVIDIA-supported project "Open Multimodal AI Infrastructure to Accelerate Science". Smith was general chair of EMNLP 2022 and program co-chair of ACL 2016.¹²³

Research

Smith's research targets systems that process language and other data to augment human capabilities, along with research methodology such as evaluation. Recent work applies multimodal AI to music, including timestamped transcription of performances (Rubato) and digitization of sheet music (Legato 2).¹²

Smith directs Ai2's OLMo open language modeling effort and co-authored the 2024 papers introducing OLMo, the Dolma pretraining corpus, the OLMoE mixture-of-experts model and the Tülu 3 post-training recipes. With Sofia Serrano and Zander Brumbaugh, Smith wrote "Language Models: A Guide for the Perplexed" (2023), a tutorial for a general audience.⁴⁵⁷⁶⁸¹

Work and contributions

• OLMo Open weights

  Directs the OLMo open language modeling effort at Ai2, per Smith's homepage; co-author of the OLMo paper (2024), and credited in OLMoE with advising on pretraining and helping frame and edit the paper.¹⁴⁷

• Tülu 3 Published research

  Co-author of "Tulu 3: Pushing Frontiers in Open Language Model Post-Training" (2024).⁶

• Dolma Open data

  Co-author of the Dolma paper, which describes a three-trillion-token English corpus released for research on language-model pretraining.⁵

Timeline

1. 2011Received an NSF CAREER award, held from 2011 to 2016.¹
2. 2016Served as program co-chair of ACL 2016.¹
3. 2017Co-led UW's Sounding Board team, which won the inaugural Amazon Alexa Prize.¹
4. 2022Served as general chair of EMNLP 2022.¹
5. Nov 2023Co-authored "Language Models: A Guide for the Perplexed", a general-audience tutorial.⁸
6. Jan 2024Co-authored the Dolma paper describing an open corpus for language model pretraining.⁵
7. Feb 2024Co-authored the OLMo paper, "Accelerating the Science of Language Models".⁴
8. Sep 2024Co-authored OLMoE, an open mixture-of-experts language model.⁷
9. Nov 2024Co-authored Tülu 3, on open post-training of language models.⁶

Honors and recognition

• Hertz Foundation Graduate Fellowship · Hertz Foundation, 2001¹
• NSF CAREER Award · U.S. National Science Foundation, 2011¹
• ACL Fellow · Association for Computational Linguistics, 2020¹
• Amazon Alexa Prize (inaugural, Sounding Board team) · Amazon, 2017¹
• University of Washington Faculty Lecture Award · University of Washington, 2026¹

Where they work

• University of Washington 19 people in the wiki
• Ai2 23 people in the wiki

Related people

• Aylin CaliskanAlso at University of Washington
• Sameer SinghAlso at University of Washington
• Yulia TsvetkovAlso at University of Washington
• Yejin ChoiAlso at University of Washington
• Ali FarhadiAlso at University of Washington
• Nicholas CarliniEvaluation and benchmarks · Language models
• Swabha SwayamdiptaEvaluation and benchmarks · Language models
• Mohit IyyerLanguage models · Evaluation and benchmarks
• Gabe PereyraEvaluation and benchmarks · Language models
• Nanyun (Violet) PengLanguage models · Evaluation and benchmarks