Beyond the Linear Representation Hypothesis: Non-Linear Activation Steering in Text-to-Image Models
Mechanistic interpretability often relies on the Linear Representation Hypothesis (LRH), which assumes that high-level concepts are encoded as linear directions in activation space.
By Muhammad Atif Butt, Pawe{\l} Skier\'s, Joost Van De Weijer, Kamil DejaarXiv - cs.CV (CV)
Read the full story
Originally published by arXiv - cs.CV (CV) on Oct 7, 2026.
Comments
Sign in to join the discussion.