This work was done during one weekend by research workshop participants and does not represent the work of Apart Research.
Accepted at the 
 research sprint on 
February 16, 2023

Automated Model Oversight Using CoTP

One aspect of scalable oversight is automated oversight, there have been some examples using models to evaluate question and model outputs, we would like to do an instantiation of this particularly using factored cognition. We’d like an automated system that is general and self-directed both with respect to inquires and topics of oversight.

Adam Khoja, Rishi Khare, John Wang
4th place
3rd place
2nd place
1st place
 by peer review