# Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills

arXiv: https://arxiv.org/abs/2506.12963 · HTML: https://mlsanity.com/paper/2506.12963/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 41 of its 48 references that are in the map.
- First posted on arXiv 15 Jun 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- LLM safety and security / [Machine unlearning](https://mlsanity.com/safety/machine-unlearning/)

## Builds on (most cited first, 30 of 41)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [Qwen2 Technical Report](https://mlsanity.com/paper/2407.10671/) (2024)
- [s1: Simple test-time scaling](https://mlsanity.com/paper/2501.19393/) (2025)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [Kimi k1.5: Scaling Reinforcement Learning with LLMs](https://mlsanity.com/paper/2501.12599/) (2025)
- [Editing Models with Task Arithmetic](https://mlsanity.com/paper/2212.04089/) (2022)
- [JailbreakBench: An Open Robustness Benchmark for Jailbreaking Large Language Models](https://mlsanity.com/paper/2404.01318/) (2024)
- [The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning](https://mlsanity.com/paper/2403.03218/) (2024)
- [Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning](https://mlsanity.com/paper/2404.05868/) (2024)
- [LIMO: Less is More for Reasoning](https://mlsanity.com/paper/2502.03387/) (2025)
- [Knowledge Unlearning for Mitigating Privacy Risks in Language Models](https://mlsanity.com/paper/2210.01504/) (2022)
- [A StrongREJECT for Empty Jailbreaks](https://mlsanity.com/paper/2402.10260/) (2024)
- [Who's Harry Potter? Approximate Unlearning in LLMs](https://mlsanity.com/paper/2310.02238/) (2023)
- [From System 1 to System 2: A Survey of Reasoning Large Language Models](https://mlsanity.com/paper/2502.17419/) (2025)
- [Large Language Model Unlearning](https://mlsanity.com/paper/2310.10683/) (2023)
- [In-Context Unlearning: Language Models as Few Shot Unlearners](https://mlsanity.com/paper/2310.07579/) (2023)
- [Reasoning Models Can Be Effective Without Thinking](https://mlsanity.com/paper/2504.09858/) (2025)
- [Simplicity Prevails: Rethinking Negative Preference Optimization for LLM Unlearning](https://mlsanity.com/paper/2410.07163/) (2024)
- [Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs](https://mlsanity.com/paper/2501.18585/) (2025)
- [SafeChain: Safety of Language Models with Long Chain-of-Thought Reasoning Capabilities](https://mlsanity.com/paper/2502.12025/) (2025)
- [LLM Post-Training: A Deep Dive into Reasoning Large Language Models](https://mlsanity.com/paper/2502.21321/) (2025)
- [Safety Tax: Safety Alignment Makes Your Large Reasoning Models Less Reasonable](https://mlsanity.com/paper/2503.00555/) (2025)
- [Large Language Model Unlearning via Embedding-Corrupted Prompts](https://mlsanity.com/paper/2406.07933/) (2024)
- [SOUL: Unlocking the Power of Second-Order Optimization for LLM Unlearning](https://mlsanity.com/paper/2404.18239/) (2024)
- [LightThinker: Thinking Step-by-Step Compression](https://mlsanity.com/paper/2502.15589/) (2025)

## Built on by (most cited first, 14 of 14)

- [A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Models](https://mlsanity.com/paper/2509.03871/) (2025)
- [Towards Safe Reasoning in Large Reasoning Models via Corrective Intervention](https://mlsanity.com/paper/2509.24393/) (2025)
- [Leak@k: Unlearning Does Not Make LLMs Forget Under Probabilistic Decoding](https://mlsanity.com/paper/2511.04934/) (2025)
- [Auditing Reasoning-Trace Memorization Claims after Unlearning with Head-Conditioned Canaries](https://mlsanity.com/paper/2605.18891/) (2026)
- [R1-ACT: Efficient Reasoning Model Safety Alignment by Activating Safety Knowledge](https://mlsanity.com/paper/2508.00324/) (2025)
- [Forgetting to Forget: Attention Sink as A Gateway for Backdooring LLM Unlearning](https://mlsanity.com/paper/2510.17021/) (2025)
- [Towards Reasoning-Preserving Unlearning in Multimodal Large Language Models](https://mlsanity.com/paper/2512.17911/) (2025)
- [Simulating Novice Students Using Machine Unlearning and Relearning in Large Language Models](https://mlsanity.com/paper/2603.26142/) (2026)
- [Unlearning as Ablation: Toward a Falsifiable Benchmark for Generative Scientific Discovery](https://mlsanity.com/paper/2508.17681/) (2025)
- [LLM Unlearning Under the Microscope: A Full-Stack View on Methods and Metrics](https://mlsanity.com/paper/2510.07626/) (2025)
- [Beyond Forgetting: Representation Misdirection Elicits Controllable Side Behaviors and Capabilities](https://mlsanity.com/paper/2601.21702/) (2026)
- [Test-Time Unlearning via Sparse Autoencoder](https://mlsanity.com/paper/2609.16229/) (2026)
- [Unlearning Deceptive Behaviors in LLMs with Contrastive Forget Sets](https://mlsanity.com/paper/2609.38909/) (2026)
- [REDaCT: Collapse-Free Unlearning in Large Reasoning Models](https://mlsanity.com/paper/2610.12769/) (2026)

## Abstract (arXiv)

Recent advances in large reasoning models (LRMs) have enabled strong chain-of-thought (CoT) generation through test-time computation. While these multi-step reasoning capabilities represent a major milestone in language model performance, they also introduce new safety risks. In this work, we present the first systematic study to revisit the problem of machine unlearning in the context of LRMs. Machine unlearning refers to the process of removing the influence of sensitive, harmful, or undesired data or knowledge from a trained model without full retraining. We show that conventional unlearning algorithms, originally designed for non-reasoning models, are inadequate for LRMs. In particular, even when final answers are successfully erased, sensitive information often persists within the intermediate reasoning steps, i.e., CoT trajectories. To address this challenge, we extend conventional unlearning and propose Reasoning-aware Representation Misdirection for Unlearning (R2MU), a novel method that effectively suppresses sensitive reasoning traces and prevents the generation of associated final answers, while preserving the model's reasoning ability. Our experiments demonstrate that R2MU significantly reduces sensitive information leakage within reasoning traces and achieves strong performance across both safety and reasoning benchmarks, evaluated on state-of-the-art models such as DeepSeek-R1-Distill-LLaMA-8B and DeepSeek-R1-Distill-Qwen-14B.