tech

December 20, 2025

Evaluating chain-of-thought monitorability

We introduce evaluations for chain-of-thought monitorability and study how it scales with test-time compute, reinforcement learning, and pretraining.

Evaluating chain-of-thought monitorability

TL;DR

  • Introduced evaluations for chain-of-thought monitorability.
  • Studied how chain-of-thought monitorability scales with test-time compute.
  • Investigated scalability with reinforcement learning.
  • Examined scalability with pretraining.

Continue reading
the original article

Made withNostr