Dominik J. Schwarz

The Entanglement Wall: Activation-Space Probes as Risk Detectors, Not Context Adjudicators

Context can change whether a request is harmful without changing its topic or surface form. We ask whether residual-stream probes distinguish harmful requests from surface-matched benign controls at a useful operating point. Across three 7-8B model families, an activation sensor blocks 95.5-97.7 percent of judge-classi...