Recent work has demonstrated that Chain-of-Thought (CoT) often yields limited gains for soft-reasoning problems such as analytical and commonsense reasoning. CoT can also be unfaithful to a model's actual reasoning. We investigate the dynamics and faithfulness of CoT in soft-reasoning tasks across instruction-tuned, reasoning and reasoning-distilled models. Our findings reveal differences in how these models rely on CoT, and show that CoT influence and faithfulness are not always aligned.
@article{arxiv.2508.19827,
title = {Analysing Chain of Thought Dynamics: Active Guidance or Unfaithful Post-hoc Rationalisation?},
author = {Samuel Lewis-Lim and Xingwei Tan and Zhixue Zhao and Nikolaos Aletras},
journal= {arXiv preprint arXiv:2508.19827},
year = {2025}
}