OpenAI Resumes Large-Scale Frontier RL Training as Astra Scores Perfect in Tests
haider1 · x · 2026-09-02
OpenAI has resumed the large-scale frontier RL (reinforcement learning) runs it had previously paused, with smaller-scale work continuing under stricter oversight, though some larger runs intended for future versions of Astra remain on hold. Separately, Astra achieved a 100% score on ExploitBench, a benchmark that tests a model's ability to build known exploits.
More from Models
- Dev complains Claude's habit of saying "not done" is his biggest pain point — jasonkneen · 2026-10-18
- Open-weight TYGER speech decision model makes 80ms turn-taking calls for full-duplex voice agents — solyarisoftware · 2026-10-18
- Paying Google AI subscriber blocked: custom-domain Workspace accounts can't access AI Studio — ChrisRackauckas · 2026-10-18
- Running Ling 3.1 Flash (560B MoE, 25B Active) Across 4 DGX Sparks via llama.cpp RPC — BerryBlush_Rogers691 · 2026-10-18
- Users press OpenAI on long-delayed Adult Mode a year after it was promised — ___Patrice___ · 2026-10-18
- Polymarket opens betting market on who releases the next Gemini Argon model — Polymarket · 2026-10-18