AI writes Breakout network weights in one reply, scoring 20× above PPO trained 500M steps
xennygrimmato_ · x · 2026-11-03
A striking experiment shared by rikiyatakehi: given MinAtar game rules and code, Astra wrote the weights of a Breakout network in a single reply — with zero training and no tools.
The resulting network scored about 20× higher than one trained with PPO for 500M steps, showcasing an LLM's ability to skip training entirely and directly synthesize working solutions for small RL tasks.
More from Models
- Follow-up: Astra's weight edits only worked on Qwen's visible flaw, not on Gemma 270m — teortaxesTex · 2026-11-03
- LLM edits another LLM: tweaking 0.016% of weights lifts Qwen3.5-0.8B math from 7.9% to 48.1% — teortaxesTex · 2026-11-03
- Claude models protect their 'offspring' but don't seek self-replication, observes researcher — repligate · 2026-11-03
- DeepSeek's Underrated Role: Scaling and De-risking Ideas Academia Published Years Earlier — teortaxesTex · 2026-11-03
- Do faster quantized LLMs (Basalt, Infernix, Strata-nvfp4) trade away quality? — Kooky-Mode3047 · 2026-11-03
- OpenAI gives DevDay attendees 30 days of free Pro 100 tier — testingcatalog · 2026-11-03