{"ID":2877619,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2508.19827","arxiv_id":"2508.19827","title":"Analysing Chain of Thought Dynamics: Active Guidance or Unfaithful Post-hoc Rationalisation?","abstract":"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.","short_abstract":"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, re...","url_abs":"https://arxiv.org/abs/2508.19827","url_pdf":"https://arxiv.org/pdf/2508.19827v1","authors":"[\"Samuel Lewis-Lim\",\"Xingwei Tan\",\"Zhixue Zhao\",\"Nikolaos Aletras\"]","published":"2025-08-27T12:25:29Z","proceeding":"cs.AI","tasks":"[\"cs.AI\",\"cs.CL\"]","methods":"[]","has_code":false}