Se­cret Loyalties

Pre-train­ing data poi­son­ing likely makes in­stal­ling se­cret loy­alties easier

How Ro­bust Is Mon­i­tor­ing Against Se­cret Loy­alties?

The Easiest Route to Se­cret Loy­alty May Be Hi­jack­ing the Model’s Chain of Command

How Se­cret Loy­alty Differs from Stan­dard Back­door Threats

Rea­son­ing Traces as a Path to Data-Effi­cient Gen­er­al­iza­tion in Data Poisoning