LLMs can natively read and write neural network weights, no training needed
Marrk · reddit · 2026-11-03
A blog post titled 'Sparks of Weight-Space Programming' shows an LLM ('Astra') directly manipulating neural network weights without tools: reading 17,253 weights to identify a Breakout-trained policy with 99% confidence, editing 3.5% of weights in one shot to lift a failed driving policy from 0% to 100% success, and editing Qwen3.5-0.8B weights via a Python kernel to boost MATH-500 accuracy from 7.9% to 48.1% with no training.
More from Models
- Leaked DeepSeek V4.1-Flash report: 552B MoE, 1M context, KV cache cut to ~1/40 — teortaxesTex · 2026-11-03
- Observers say Dessa eerily matches Karpathy's "summoning ghosts" analogy; Jack calls it fast — analisereal · 2026-11-03
- Together at TC Disrupt: 'token maxing without cost maxing' via smaller models — togethercompute · 2026-11-03
- Dream-1 zero-shot robot test: 100 public episodes show real skills but poor instruction following — paigeinsf · 2026-11-03
- Speculation: DeepSeek to ship V4.1F or V4.2F before any Pro variant — teortaxesTex · 2026-11-03
- Follow-up: Astra's weight edits only worked on Qwen's visible flaw, not on Gemma 270m — teortaxesTex · 2026-11-03