OpenAI's Astra uses recurrent depth to hide reasoning, but experts say it's a minor tweak
dotey · x · 2026-09-06
The Information reports OpenAI's Astra uses "recurrent depth" (looped transformer), letting the model repeatedly process the same text for higher quality while hiding part or all of its chain of thought, making internal reasoning hard to inspect.
- Insiders say OpenAI deliberately limited the technique so Astra still produces human-readable CoT for monitoring; the official blog also promised "additional chain-of-thought monitoring" at launch.
- Industry researchers, including the U.K. AI Security Institute, worry other developers adopting the technique without such restraint could produce unmonitorable "runaway AI."
- Architectures expert Sebastian Raschka counters that looped transformers aren't new nor very important: "Astra may be a great model, but it likely has little to do with it — just a minor architectural tweak."
Related event: OpenAI's Astra Reportedly Uses Recurrent Depth, Raising AI Safety Concerns(7 posts)→
More from Models
- Dev review: Astra 'feels like Sol 5.7 on a good day', unusable for security work — jarrodwatts · 2026-09-11
- GPT-6 Astra docs draw attention: async tool calling and mid-turn steering point to multi-agent use — MikkoH · 2026-09-11
- Leaked GPT-6 Astra Scores 46% vs 12% for MolmoAct2 on Bimanual Robot Tasks — damianplayer · 2026-09-11
- OpenAI ships GPT-Live prompting guide: copying your old prompts won't hit SOTA — craigsdennis · 2026-09-11
- GPT-6 Astra burns through user's weekly usage cap, forcing a wait until Monday — max_paperclips · 2026-09-11
- Anthropic blocks minors from using Claude, HN debates age policy — petrusenko_max · 2026-09-11