SOURCE-LINKED INTELLIGENCE

Decodability is Not Causality: Dissociating Probe Readouts from Behavioral Drivers via SAE Decomposition

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

Linear probes can decode safety-relevant concepts such as truthfulness from language-model activations, but probe accuracy may show only decodability, not that the features the probe weights causally drive model behavior. We demonstrate that this gap cannot be closed from the geometry of probe weights alone: the features geometrically aligned with probe direction need not be the ones the model uses, so causal relevance requires intervention. We introduce a feature-level diagnostic that decomposes a deployed True/False probe into sparse-autoencoder (SAE) features, ranks those features by both p

Read original source ↗ Open in workspace

recordType

paper

region

Global