Auditable AI-text detectors with sparse autoencoders: detectors, heads, descriptions and Gemma-2-2b extractors.
-
Detector Scope
🔬Auditable AI-text detection with sparse-autoencoder features
-
Building Auditable AI Text Detectors with Sparse Autoencoders
Paper • 2610.30006 • Published -
liamdugan/detector-scope
Text Classification • 0.3B • Updated • 318 -
liamdugan/detector-scope-peer-review-gpt4o
Text Classification • 0.3B • Updated • 36
