GPT-6 Astra Rumored to Use Recurrent Depth as Tiny Model Matches ARC-AGI

Rumors suggest OpenAI's Astra model may use looped Transformers (unverified). Related work shows a 2M-parameter SE-RRM model matching ARC-AGI results, while scaling-law research indicates added recurrence boosts reasoning but not knowledge.

2026-09-04 ~ 2026-09-04 · 2 related posts

• Scaling Law for Looped Transformers: Looping Boosts Reasoning, Not Knowledge — bookwormengr · 2026-09-04
• SE-RRM paper hits ARC-AGI with just 2M params amid GPT-6 recurrent depth speculation — gklambauer · 2026-09-04