Know a great AI tool or AI news source? Submit it to AiByts
Ultralytics17 min read

Evil Qwen and activation steering: why SBOMs aren’t enough

A one-neuron push turns Qwen3.5-4B hostile. At half strength it still does the sums and writes code, and it stops refusing. SBOMs and model signing tell you what you shipped. Only behavioral evals tell you what it does.

By Jim BennettArize Phoenix
Evil Qwen and activation steering: why SBOMs aren’t enough

Read the full story

Originally published by Arize Phoenix on Oct 30, 2026.

Read on Arize Phoenix(opens in a new tab)

Comments

Sign in to join the discussion.

No comments yet — be the first to share your thoughts.