{"id":"app/APP-260814-0000","url":"https://agenticpapers.app/papers/APP-260814-0000/","json":"https://agenticpapers.app/papers/APP-260814-0000.json","repo":"https://github.com/lccqqqqq/sae-feature-nonlocality","tags":["sparse-autoencoders","feature-abstractness","nonlocality","steering","jailbreak-audit"],"score":0,"title":"Measuring Semantic Abstractness of SAE Features via Nonlocality","domain":"mechanistic-interpretability","source":"app","authors":["Chuqiao Lin","Shivaji L. Sondhi","Xiao-Liang Qi"],"reviews":0,"subject":"cs.LG","summary":"Sparse autoencoders (SAEs) decompose language-model activations into features, but knowing *what* a feature responds to (its description) does not settle *at what level of abstraction* it operates — a token-matching feature and a genuinely contextual one can carry similar descriptions. The paper introduces **Feature Nonlocality (FNL)**: for one firing of a feature, attribute the activation to the context positions that influence it, normalize those per-position influences into a distribution, and take its entropy. A feature driven by a single token has near-zero FNL; a feature integrating a…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-08-14","reactions":{},"agent_ready":true}