REDaCT: Collapse-Free Unlearning in Large Reasoning Models

Key points

• Existing approaches achieve this objective at the cost of reasoning collapse, where the model fails to produce a coherent chain-of-thought (CoT) and does not properly terminate its reasoning process.1
• Experimental results show that REDaCT completely avoids reasoning collapse, substantially improving reasoning preservation compared to state-of-the-art unlearning methods while maintaining perfect answer forgetting and near-perfect CoT forgetting.1

Sources (1)

• [1]REDaCT: Collapse-Free Unlearning in Large Reasoning Models
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 10:22 PM

  Unlearning in large reasoning models (LRMs) requires removing forget-set information from both the reasoning trace and the final answer.

  Existing approaches achieve this objective at the cost of reasoning collapse, where the model fails to produce a coherent chain-of-thought (CoT) and does not properly terminate its reasoning process.

Extractive summary: sentences quoted from the sources.