SOURCE-LINKED INTELLIGENCE

The Oversight Gap: What LLM Safety Monitors Miss, and Why It Is Not Capability

arXiv · AI, language, vision and robotics · article · Sep 7, 2026 · UTC

Several properties safety monitors are asked to certify, among them cross-tenant noninterference, sandbagging and evaluation awareness, are 2-safety hyperproperties, witnessed only by two executions. The standard consequence is a binary impossibility: one trace cannot decide them. We replace the binary with a measurement. A tight bound puts the balanced accuracy of any single-trace monitor at $\tfrac12+\tfrac12\,TV(P_0,P_1)$, turning undecidability into a graded detectability frontier and defining an oversight gap: a monitor's shortfall below it. On a leak family with closed-form $TV$, nine LL

Read original source ↗ Open in workspace

recordType

paper

region

Global