[advik basani]
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

> blog posts

Thoughts on machine learning, research, and technology

--------------------------------------------------------------------------------
BLOGJune 2026LessWrong

Brittle Model Organisms Obstructs Deception Elicitation Work

Advik Raj Basani, Daniel Tan, Chloe Li

Finetuning-based auditing methods meant to measure deceptive behavior in AI model organisms inadvertently destroy that behavior while eliciting false confessions, undermining the validity of these evaluation approaches.

--------------------------------------------------------------------------------