Externalizing Reasoning

Interpretabilityemerging

Training models to reason through extended, readable chains of thought rather than opaque internal computation.

Cluster: Interpretability

Tags