Allen Institute for AI

Last updated
Allen Institute for AI
Formation2014;11 years ago (2014)
Founder Paul Allen
Type Non-profit research institute
82-4083177
Location
CEO
Ali Farhadi
Key people
Peter Clark, Yejin Choi, Noah Smith, Hannaneh Hajishirzi, Dan Weld, Chris Bretherton, Ani Kembhavi, Jes Lefcourt
Website allenai.org OOjs UI icon edit-ltr-progressive.svg

The Allen Institute for AI (abbreviated AI2) is a 501(c)(3) non-profit scientific research institute founded by late Microsoft co-founder and philanthropist Paul Allen in 2014. The institute seeks to conduct high-impact AI research and engineering in service of the common good. [1] AI2 is based in Seattle, and also has an active office in Tel Aviv, Israel. [2]

Contents

History

Oren Etzioni was appointed by Paul Allen in September 2013 to direct the research at the institute. [3] After leading the organization for nine years, Etzioni stepped down from his role as CEO on September 30, 2022. [4] He was replaced in an interim capacity by the leading researcher of the company's Aristo project, Peter Clark. [5] On June 20, 2023, AI2 announced Ali Farhadi as its next CEO starting July 31, 2023. [5]

Teams

Generative AI

OLMo model family

On May 11, 2023, AI2 announced they were developing OLMo, an open language model aiming to match the performance of other state-of-the-art language models. [14] In February 2024, 1B and 7B parameter variants of the model were open-sourced, including code, model weights with intermediate snapshots and logs, and contents of their Dolma training dataset, [15] making it the most open state-of-the-art model available. [16] [17] In November 2024, AI2 released the second iteration of OLMo, OLMo 2, with the initial release including 7B and 13B parameter models. [18] [19] In March 2025, AI2 released a 32B variant of OLMo 2, claiming to have released "the first fully-open model (all data, code, weights, and details are freely available) to outperform GPT3.5-Turbo and GPT-4o mini". [20]

Tulu models and post-training recipes

In addition to the fully-open OLMo family of models, AI2 has also developed Tulu, a family of instruction-tuned models and open post-training recipes that build on open-weights base models (e.g., Meta's Llama) to provide fully transparent alternatives to proprietary instruction-tuning methods. [21] AI2 released the first iteration of Tulu in June 2023, [22] with subsequent iterations being released in November 2023 (Tulu 2 [23] ) and November 2024 (Tulu 3 [24] ).

See also

References

  1. "About — Allen Institute for AI". allenai.org. Retrieved 2023-05-26.
  2. "AI2 Israel — Allen Institute for AI". allenai.org. Retrieved 2023-05-26.
  3. Cook, John (2013-09-04). "Going beyond Siri and Watson: Microsoft co-founder Paul Allen taps Oren Etzioni to lead new Artificial Intelligence Institute". GeekWire. Retrieved 2023-05-26.
  4. Schlosser, Kurt (2022-06-15). "Oren Etzioni stepping down as CEO of Allen Institute for AI after nine years at research hub". GeekWire. Retrieved 2023-05-26.
  5. 1 2 Bishop, Todd (2023-06-20). "Apple machine learning leader Ali Farhadi named CEO of Allen Institute for Artificial Intelligence". GeekWire . Archived from the original on 2023-06-20.
  6. "Aristo — Allen Institute for AI". allenai.org. Retrieved 2023-05-26.
  7. "Aristo — Allen Institute for AI". allenai.org. Retrieved 2023-05-26.
  8. "PRIOR". prior.allenai.org. Retrieved 2023-05-26.
  9. "AI2-THOR". Allen Institute for AI. Retrieved 2023-05-26.
  10. Rodriguez, Jesus (2021-07-08). "🔹🔸Edge#104: AllenNLP Makes Cutting-Edge NLP Models Look Easy". TheSequence. Retrieved 2023-05-26.
  11. "Semantic Scholar | Product". www.semanticscholar.org. Retrieved 2023-05-26.
  12. "AllenNLP — Allen Institute for AI". allenai.org. Retrieved 2023-05-26.
  13. Dormehl, Luke (2018-04-13). "Forget Cloning, A.I. is the Real Way to Let Your Family Pooch Live Forever". Digital Trends. Retrieved 2023-05-26.
  14. Groeneveld, Dirk; Beltagy, Iz; Walsh, Pete; Bhagia, Akshita; Kinney, Rodney; Tafjord, Oyvind; Jha, Ananya Harsh; Ivison, Hamish; Magnusson, Ian (2024), OLMo: Accelerating the Science of Language Models, arXiv, doi:10.48550/ARXIV.2402.00838 , retrieved 2025-08-26
  15. Soldaini, Luca; Kinney, Rodney; Bhagia, Akshita; Schwenk, Dustin; Atkinson, David; Authur, Russell; Bogin, Ben; Chandu, Khyathi; Dumas, Jennifer (2024), Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research, arXiv, doi:10.48550/ARXIV.2402.00159 , retrieved 2025-08-26
  16. AI2 (2023-05-18). "Announcing AI2 OLMo, an open language model made by scientists, for scientists". Medium. Retrieved 2023-05-26.{{cite web}}: CS1 maint: numeric names: authors list (link)
  17. Wiggers, Kyle (2024-02-01). "AI2 open sources text-generating AI models -- and the data used to train them". TechCrunch. Retrieved 2024-02-03.
  18. "OLMo 2: The best fully open language model to date | Ai2". allenai.org. Retrieved 2025-08-26.
  19. OLMo, Team; Walsh, Pete; Soldaini, Luca; Groeneveld, Dirk; Lo, Kyle; Arora, Shane; Bhagia, Akshita; Gu, Yuling; Huang, Shengyi (2025), 2 OLMo 2 Furious, arXiv, doi:10.48550/ARXIV.2501.00656 , retrieved 2025-08-26
  20. "OLMo 2 32B: First fully open model to outperform GPT 3.5 and GPT 4o mini | Ai2". allenai.org. Retrieved 2025-08-26.
  21. Coldewey, Devin (2024-11-21). "Ai2's open source Tülu 3 lets anyone play the AI post-training game". TechCrunch. Retrieved 2025-08-26.
  22. Wang, Yizhong; Ivison, Hamish; Dasigi, Pradeep; Hessel, Jack; Khot, Tushar; Chandu, Khyathi Raghavi; Wadden, David; MacMillan, Kelsey; Smith, Noah A. (2023), How Far Can Camels Go? Exploring the State of Instruction Tuning on Open Resources, arXiv, doi:10.48550/ARXIV.2306.04751 , retrieved 2025-08-26
  23. Ivison, Hamish; Wang, Yizhong; Pyatkin, Valentina; Lambert, Nathan; Peters, Matthew; Dasigi, Pradeep; Jang, Joel; Wadden, David; Smith, Noah A. (2023), Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2, arXiv, doi:10.48550/ARXIV.2311.10702 , retrieved 2025-08-26
  24. Lambert, Nathan; Morrison, Jacob; Pyatkin, Valentina; Huang, Shengyi; Ivison, Hamish; Brahman, Faeze; Miranda, Lester James V.; Liu, Alisa; Dziri, Nouha (2025-04-14), Tulu 3: Pushing Frontiers in Open Language Model Post-Training, arXiv, doi:10.48550/arXiv.2411.15124, arXiv:2411.15124, retrieved 2025-08-26