x

LESSWRONG
LW

Yueh Han Chen — LessWrong

Yueh Han Chen

Yueh Han Chen

Message

16

6mo

Yueh Han Chen hasn't written anything yet.

Yueh Han Chen

16

6mo

Yueh Han Chen has not written any posts yet.

Training Agents to Self-Report Misbehavior

TL;DR: Frontier AI agents may pursue hidden goals while concealing this pursuit from oversight. Currently, we use two main approaches to reduce this risk: (1) Alignment trains the agent to not misbehave, (2) Blackbox monitoring uses a separate model to detect misbehavior. We study a third approach—self-incrimination—which trains agents to...