Neuron steering struggles at persona creation, failing multi-turn stability
davidad · x · 2026-10-29
Developer codyhxyz shared early experiment results: neuron steering isn't particularly helpful for persona creation, with the resulting personas failing at multi-turn stability. A blogpost may be coming, but the author is largely back to the drawing board.
davidad added that he'd bet this type of alteration even makes Qwen3 8B/27B weaker than default at making plans that actually work, especially multi-agent plans.
Related event: Neuron Steering Fails to Create Stable Model Personas(2 posts)→
More from Research
- Ben Recht Debunks 'Silicon Sampling': Why Using LLMs as Survey Respondents Is Confusing — beenwrekt · 2026-10-29
- Mila's open-source Harpia uses drones + AI to survey hundreds of tree crowns in a day — Mila_Quebec · 2026-10-29
- Sakana AI takes 1st place in OpenMath 2026 math hackathon with 14 mathematical families, 6 full resolutions — SakanaAILabs · 2026-10-29
- Helm activation steering: only 10 of 53 pushes held, author shares full data — leolambertini · 2026-10-29
- Ofir Press: Millennium Problems is AI's best benchmark; 70% of open math could fall in a year — OfirPress · 2026-10-29
- Meta and Oxford's OverSeer lifts multi-agent world-model consistency on Minecraft to 60.2% from 0–12.5% baselines — armand_ruiz · 2026-10-29