
Independent company · San Francisco, United States
Pluralis Research
A research lab developing Protocol Learning, where large models are trained across many untrusted participants without any one holding the full weights.
San Francisco, United States
- Founded
- 2024
- Type
- Independent company
- Based in
- San Francisco, United States
- Key papers
- 1
- Featured projects
- 4
About
Pluralis Research was founded by Alexander Long, an ML PhD who previously worked in Amazon's international machine learning group. The lab is backed by investors including Union Square Ventures and CoinFund.
Its foundational paper, Protocol Learning, Decentralized Frontier Risk and the No-Off Problem, appeared in December 2024. The lab then published work on model-parallel training over low-bandwidth links, including Subspace Networks and Unextractable Protocol Models at NeurIPS 2025.
Code libraries are on GitHub, and its Agora technical report describes internet-scale pretraining across permissionless participants.
Mission and ambitionsPluralis argues that frontier models should be trainable by open networks rather than by a few firms, and works on the compression and parallelism needed.
Known for
- 01
Protocol Learning concept
- 02
Low-bandwidth model-parallel training
- 03
Unextractable protocol models
Milestones
- 2024
Protocol Learning paper
Long posted the position paper on arXiv in December.
- 2025
Nesterov method for asynchronous pipeline parallel optimization
Paper accepted at ICML 2025.
- 2025
Subspace Networks
Communication-efficient model parallelism paper at NeurIPS 2025.
- 2025
Unextractable Protocol Models
Paper on training and inference without weight materialization, NeurIPS 2025.
- 2026
Agora technical report
Report on collective, permissionless pretraining of LLMs.
Open source
Node0.
Collaborative training library.
Async pipeline parallel.
RL post-training.
Research papers
Notable releases
Openness
Open code on GitHub; no public weights seen.
People
- Alexander LongFounder
- Thalaiyasingam AjanthanResearcher, paper author
- Sameera RamasingheResearcher, paper author
Listed from the lab's own pages and public profiles.
Contact and community
Contact details and links are on the lab's website (https://pluralis.ai); code questions go through issues on their GitHub organisation.
Sources
Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.



