causal
Internal training-check defeats the deceiver
A deceptive model that only misbehaves when it believes it's not being watched is actually catchable, because the explanation for its good training-time behavior must factor through its internal 'am I being trained?' check; once that check is an essential part of the explanation, a deployment input where the check fails makes the explanation break down completely, flagging the relevant anomaly even if the output looks the same.
causalpending
Speaker
Paul ChristianoEvidence Quote
“if you're like why did you not stab people on training time?... it's just constantly thinks that it's being trained. Like, this thing is always this nonlinear feature.”
Created: 6/13/2026, 3:36:05 AM
Your notes
Loading notes...