$1,000-trained HRM-Text shows Sapient bet on recurrence before OpenAI's Astra
rohanpaul_ai · x · 2026-09-06
- Sapient introduced its Hierarchical Reasoning Model (HRM) in June 2025 and open-sourced HRM-Text in May.
- The core idea: reasoning can spend compute by repeatedly updating internal state via recurrent loops, rather than relying only on a conventional transformer stack — the same direction OpenAI's GPT-6 Astra is now pushing into the frontier conversation.
- HRM-Text is a 1B-parameter open-source model pretrained from scratch on just 40B unique tokens for roughly $1,000 in GPU compute, with a full pretraining framework released for others to inspect, retrain, and build on.
Related event: Sapient's Open-Source HRM-Text Model Tops 250K Downloads(2 posts)→
More from Models
- Meme: Google may deem Gemini 4 Pro not worth it and pivot back to Flash models — Able-Line2683 · 2026-09-07
- Astra robot-control demos pile up; speculation OpenAI is building its own Gemini Robotics rival — CyberRobooo · 2026-09-07
- Professor says Astra drafted five NIH R01 grant proposals in an hour for ~$20 of compute — mmbronstein · 2026-09-07
- GPT-6 "Astra" shows zero-shot robot arm sorting; OpenAI robotics rumors swirl — CyberRobooo · 2026-09-07
- Users notice GPT-6's aggressive length penalty producing unusually condensed code — cneuralnetwork · 2026-09-07
- GPT-6 Astra less than 4 days old, KOL shares 10 wild demos as users call it AGI-level — minchoi · 2026-09-07