Tim Bakker
Tim Bakker
Home
Talks
Publications
Blog
Supervision
Teaching
Experience
Organiser
Contact
Light
Dark
Automatic
Chain-of-thought
Analyzing and Improving Chain-of-Thought Monitorability Through Information Theory
We use information theory to analyze and improve chain-of-thought monitorability, proposing training methods that improve monitor accuracy while preventing CoT degeneration.
Cite
×