Published on February 15, 2026 5:49 PM GMTWe introduce an automated activation‑steering approach that plugs into standard labeled datasets—no handcrafted prompt pairs or feature annotation. On 18 tasks and 3 open‑weight models, the introspective variant (iPAS) yields the strongest behavior improvements, and layers on top of ICL/SFT.Full write‑up: https://open.substack.com/pub/sashacui/p/painless-activation-steering-pasPaper: arxiv.org/abs/2509.22739 Discuss Read More
Painless Activation Steering
Published on February 15, 2026 5:49 PM GMTWe introduce an automated activation‑steering approach that plugs into standard labeled datasets—no handcrafted prompt pairs or feature annotation. On 18 tasks and 3 open‑weight models, the introspective variant (iPAS) yields the strongest behavior improvements, and layers on top of ICL/SFT.Full write‑up: https://open.substack.com/pub/sashacui/p/painless-activation-steering-pasPaper: arxiv.org/abs/2509.22739 Discuss Read More
Related Posts
Vibestemics
Published on February 4, 2026 4:40 PM GMTA few months ago I coined the word…
Demands Are All You Need: Prompt Imperativeness Drastically Reduces Hedging In LLMs (n=900, Cohen’s d = 2.67)
Published on February 6, 2026 1:22 PM GMTForewordIf you've ever noticed LLMs hedging considerably more…
An Opinionated Guide to Privacy Despite Authoritarianism
Published on October 29, 2025 8:32 PM GMTI've created a highly specific and actionable privacy…