x

LESSWRONG

LW

Sebastian Prasanna — LessWrong

Sebastian Prasanna

Sebastian Prasanna

Message

31

1

3mo

Sebastian Prasanna

31

3mo

Model organisms researchers should check whether high LRs defeat their model organisms

Thanks to Buck Shlegeris for feedback on a draft of this post. The goal-guarding hypothesis states that schemers will be able to preserve their goals during training by taking actions which are selected for by the training process. To investigate the goal-guarding hypothesis, we’ve been running experiments of the following...

Basic Legibility Protocols Improve Trusted Monitoring

This is a blog post for our paper Basic Legibility Protocols Improve Trusted Monitoring. This work was done as part of the CBAI Summer Research Fellowship, mentored by Cody Rushing. In this blog post, we’ll go over the main results, and focus more on the engineering and methodological takeaways than...