MIT Weight-Space Programming Independently Replicated, with 20x Gains on Newer Models
Alex Krentsel independently replicated MIT's weight-space programming results, finding that frontier models can directly read and edit neural network weights without gradient training. The capability appears only in the newest models such as GPT-6 Astra and Opus 5.5, showing roughly 20x gains, though some question calling it 'emergent'.
2026-11-03 ~ 2026-11-03 · 4 related posts
- Episode 1: MIT: LLM Edits 0.016% of Weights to Boost Small Model Math 6x(2026-11-03, 47 posts)
- Episode 2: Researchers Clarify Astra Weight Editing: Known Interpretability Tech, Not RSI(2026-11-03, 3 posts)
- Episode 3: GPT-6 'Weight Editing' Claim Disputed as Steering Vector Trial-and-Error(2026-11-03, 2 posts)
- Episode 4: MIT Weight-Space Programming Independently Replicated, with 20x Gains on Newer Models(2026-11-03, 4 posts)
- MIT weight-space programming independently reproduced; GPT-6 Astra hits 22x, Opus 5.5 20x — a1zhang · 2026-11-03
- Weight-space programming result reproduced, but 'emergent' claim disputed — anshulkundaje · 2026-11-03
- MIT weight-space programming results reproduced; step-change capability in newest models — basedjensen · 2026-11-03
- Weight-space programming emerges as a new capability, 20x jump in latest models — Distinct-Question-16 · 2026-11-03