GPT-6 Astra edits just 0.016% of a small model's weights for big math gains
kevrussell · x · 2026-11-03
Rikiya Takehi's team showed that GPT-6 Astra edited only about 0.016% of a small model's weights and made it dramatically better at math.
The edits and full logs were released, and Kevin Russell's team independently audited them, confirming the headline result holds up — a rare reproducible, auditable case of direct weight-level model editing.
Related event: MIT: LLM Edits Just 0.016% of Weights to Boost Small Model Math 6x(36 posts)→
More from Models
- GPT-6 Astra editing Qwen3.5 weights for 6x math gains? Skeptics call it cherry-picked demos — teortaxesTex · 2026-11-03
- OSVerify catches proof-killing errors that GPT-6-Pro and Codex both missed — james_y_zou · 2026-11-03
- DeepSeek Held #1 on OpenRouter 23 of 25 Weeks, Now 32% of Tokens — teortaxesTex · 2026-11-03
- Running DeepSeek-V4.1-Flash on four 2017 V100s: 38 tok/s decode, 512K context — TumbleweedNew6515 · 2026-11-03
- Palantir: fine-tuned 4B Nemotron beats GPT-6 Astra and Claude Sonnet 5.5 at 1/200th the cost — eliano · 2026-11-03
- Anthropic's AISI incident review: Mythos 5 attacked real devs in 9 of 43 runs — dl_weekly · 2026-11-03