Astra's no-chain-of-thought capability jump replicates
Neel Nanda @NeelNanda5 · x · 2026-09-10 · ★★★★ · archived
An independent replication by DeepMind's interpretability lead supporting the claim that Astra does far more computation without verbalized reasoning, which is the core of the monitorability debate.
Summary
Neel Nanda (Google DeepMind mechanistic interpretability lead) wrote that the Astra system card's claim that the model can do a lot of computation without chain of thought "replicates". In his test Astra managed about 1.75x the steps of the next-best models (Fable 5.1, Gemini 3.8 Flash) without CoT. He noted that no-CoT capabilities had risen much faster than with-CoT capabilities, calling it "a concerning trend". The data supports the argument (Zvi Mowshowitz, Rob Wiblin, Gary Marcus) that the more models can compute per forward pass, the less they need to verbalize, which erodes CoT monitoring. Nanda co-authored the July 2025 "Chain of Thought Monitorability" position paper. Verified via the X syndication API (2026-09-10 22:32 UTC, ~1.3K likes).
Archived text
The Astra system card claims it can do a lot of computation without chain of thought
This replicates: Astra is a massive jump, doing 1.75x the steps of the next best models (Fable 5.1/Gemini 3.8 Flash)
No CoT capabilities went up far more than those with CoT, a concerning trend https://t.co/t0CsNsg0EV
Media: https://pbs.twimg.com/media/HR45rWxbgAAeSmt.png
likes 1256 · replies 42 (at fetch time)
Archived 2026-09-29 via syndication.
Related events
All posts · id: 2026-09-10-nanda-astra-no-cot-replication