Valency
alpha

Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models

Carson DenisonMonte MacDiarmidFazl BarezDavid DuvenaudShauna Kravec
Fresh — published within the last 30 daysv1 publishedLicenseRef-arXiv-nonexclusive-distrib-1.0
AI contributions — not recorded for this paperView details
Machine checks

No machine checks are available from the Hub API for this paper.