Do explanations generalize across large reasoning models?

Koyena Pal, David Bau, Chandan Singh

Large reasoning models (LRMs) produce a textual chain of thought (CoT) in the process of solving a problem, which serves as a potentially powerful tool to understand the problem by…