AI CAPITAL INTELLIGENCE
Nathan Lambert | Profile
Profile at a glance
Nathan Lambert is a machine-learning researcher whose work focuses on reinforcement learning, post-training and open language models. His September 2026 CV records research roles at Ai2 (2023–2026), Hugging Face (2022–2023), DeepMind and Facebook AI, plus a PhD from UC Berkeley and a bachelor’s degree from Cornell. His current personal bio describes an unnamed nonprofit open-science initiative and the Interconnects AI newsletter; it does not identify the organization.
Source verification date: · Career dates and degrees use Nathan Lambert’s CV, last updated September 14, 2026. His personal bio names no organization for the nonprofit open-science effort, so no name is inferred. Research contributions and awards are attributed to his CV.
Names follow their disclosed form. Descriptive English renderings of Chinese names are not verified legal names. Original quotations are shown in their source language.
Key facts
| Research focus | Reinforcement learning, post-training and open language models |
|---|---|
| Current public work | Co-founder of an unnamed nonprofit open-science initiative; writer of Interconnects AI |
| Recent research roles | Allen Institute for AI, 2023–2026; Hugging Face, 2022–2023 |
| Education | PhD, UC Berkeley; BS, Cornell University |
Experience
Lambert’s career spans robotics and reinforcement-learning research, human-feedback methods, open language models and research communication. His CV lists roles at Ai2 and Hugging Face; his personal bio describes an open-science nonprofit without naming it.
Career and roles
- Interconnects AI · Writer
Current role (as reported in the cited source)
- Allen Institute for AI · Senior Research Scientist and post-training lead
2023–2026 · His personal bio says he co-led the teams building OLMo 1, 2 and 3 and Tülu 2 and 3.
- Hugging Face · Research Scientist and RLHF Team Lead
2022–2023 · Helped build an RLHF research team.
- DeepMind · Research Intern
2021
- Facebook AI · Research Intern and Student Researcher
2019–2020
- Tesla · Test Engineering Intern
2015
- Undisclosed nonprofit initiative · Co-founder; name not disclosed in public bio
Current self-description verified September 27, 2026 · The bio describes an effort focused on open-science methods for frontier AI.
Education
- University of California, Berkeley
PhD in Electrical Engineering and Computer Science, 2017–2022
- Cornell University
BS in Electrical and Computer Engineering, 2013–2017
Milestones
- 2026
His September 2026 CV lists his Allen Institute for AI research role through 2026; his current biography describes a new nonprofit open-science initiative.
Public record
- His CV lists post-training work on OLMo 1, 2 and 3 and Tülu 2 and 3, plus a 2025 CVPR Best Paper Honorable Mention for Molmo & Pixmo and 2024 ACL awards for Dolma and OLMo.
- He writes Interconnects AI, a newsletter about open models and AI research, according to his current personal bio.
Sources
- 1. CV · Nathan Lambert
Allen Institute for AI, Seattle, WA — Research Scientist — 2023 – cont. HuggingFace, Remote — Research Scientist and RLHF Team Lead — 2022 – 2023. DeepMind, London (Virtual) — Research Intern — 2021. Ph.D. in Electrical Engineering and Computer Science, University of California, Berkeley — 2017 – 2022. B.S. in Electrical and Computer Engineering, Cornell University — 2013 – 2017.
- 2. Contact Nathan Lambert · Nathan Lambert
Nathan Lambert is a co-founder and {stealth} of {stealth}, a non-profit with the mission of fostering the open science of frontier AI, which currently focuses on building open post-training recipes, studying how models get their character, and understanding the open-weight model ecosystem. He also writes Interconnects AI, a well read technical newsletter on open models and AI research.
- 3. CV · Nathan Lambert
Allen Institute for AI, Seattle, WA — Research Scientist 2023 – 2026. HuggingFace, Remote — Research Scientist and RLHF Team Lead 2022 – 2023. DeepMind, London (Virtual) — Research Intern (Host: Martin Riedmiller) 2021. Facebook AI, Menlo Park — Research Intern and Student Researcher (Host: Roberto Calandra) 2019 – 2020. Tesla, Palo Alto Test Engineering Intern 2015. Ph.D. in Electrical Engineering and Computer Science, University of California, Berkeley (4.0/4.0) — 2017 – 2022. B.S. in Electrical and Computer Engineering, Cornell University (4.0/4.0) — 2013 – 2017. CVPR Best Paper Honorable Mention (Molmo & Pixmo) 2025; ACL Best Resource Paper Award (Dolma) 2024; ACL Best Theme Paper Award (OLMo) 2024.