Year: 2025
Type: preprint
Abstract: We present INTELLECT-3, a 106B-parameter Mixture-of-Experts model (12B active) trained with large-scale reinforcement learning on our end-to-end RL infrastructure stack. INTELLECT-3 achieves state of the art performance for its size across math, code, science and reasoning benchmarks, outperforming many larger frontier models. We open-source the model together with the full infrastructure stack used to create it, including RL frameworks, complete recipe, and a wide collection of environments, built with the verifiers library, for training and evaluation from our Environments Hub community platform. Built for this effort, we introduce prime-rl, an open framework for large-scale asynchronous reinforcement learning, which scales seamlessly from a single node to thousands of GPUs, and is tailored for agentic RL with first-class support for multi-turn interactions and tool use. Using this stack, we run both SFT and RL training on top of the GLM-4.5-Air-Base model, scaling RL training up to 512 H200s with high training efficiency.
Source: arXiv (Cornell University)
Authors: Prime Intellect Team, Senghaas, Mika, Obeid, Fares, Jaghouar, Sami, Brown, William, Ong, Jack Min, Auras, Daniel, Sirovatka, Matej, Straube, Jannik, Baker, Andrew, Müller, Sebastian, Mattern, Justus, Basra, Manveer, Ismail, Aiman, Scherm, Dominik, Miller, Cooper, Patel, Ameen, Kirsten, Simon, Sieg, Mario, Reetz, Christian, Erdem, Kemal, Weisser, Vincent, Hagemann, Johannes
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4417529825
Cites:
Cited by:
Related to: