Preference optimization can learn spurious correlations, and augmenting the preference data with ties (equally preferred responses) can reduce spurious reliance. However, ties differ in their corrective effect, making their selection for augmentation important. For log-linear policies, we derive each tie's influence on...
Christian Moya, Purav Matlia, E. Thornley et al.· 0 citations
Theory and experiments suggest that capable AI agents may resist shutdown. Existing shutdown evaluations are informative but limited in coverage and vulnerable to evaluation awareness. We introduce ShutdownMod, a method that turns any agent trajectory into a shutdown evaluation by injecting a shutdown request into the...
Param Biyani, Sebastien Luyasu, Cedric Potvliege et al.· 0 citations
It is shown that selective control under partial auditing reduces accepted errors while increasing correctness, and experiments with log linear and neural contextual bandits and with a language model support the analysis and show that selective control under partial auditing reduces accepted errors while increasing cor...
Christian Moya, E. Thornley, Guang Lin· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.