# Towards Reasoning-Preserving Unlearning in Multimodal Large Language Models

arXiv: https://arxiv.org/abs/2512.17911 · HTML: https://mlsanity.com/paper/2512.17911/

## Key facts

- Cited by 4 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 40 of its 84 references that are in the map.
- First posted on arXiv 26 Nov 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- Multimodal LLMs / vision-language models / [MLLM jailbreaks and adversarial robustness](https://mlsanity.com/mllm/mllm-jailbreaks-and-adversarial-robustness/)
- LLM safety and security / [Machine unlearning](https://mlsanity.com/safety/machine-unlearning/)

## Builds on (most cited first, 30 of 40)

- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [Negative Preference Optimization: From Catastrophic Collapse to Effective Unlearning](https://mlsanity.com/paper/2404.05868/) (2024)
- [Chain-of-Note: Enhancing Robustness in Retrieval-Augmented Language Models](https://mlsanity.com/paper/2311.09210/) (2023)
- [Recent Advances in Optimal Transport for Machine Learning](https://mlsanity.com/paper/2306.16156/) (2023)
- [Reducing Hallucinations in Vision-Language Models via Latent Space Steering](https://mlsanity.com/paper/2410.15778/) (2024)
- [Locate-then-edit for Multi-hop Factual Recall under Knowledge Editing](https://mlsanity.com/paper/2410.06331/) (2024)
- [Evaluating and Steering Modality Preferences in Multimodal Large Language Model](https://mlsanity.com/paper/2505.20977/) (2025)
- [MMUnlearner: Reformulating Multimodal Machine Unlearning in the Era of Multimodal Large Language Models](https://mlsanity.com/paper/2502.11051/) (2025)
- [Dissecting Fine-Tuning Unlearning in Large Language Models](https://mlsanity.com/paper/2410.06606/) (2024)
- [Modality-Aware Neuron Pruning for Unlearning in Multimodal Large Language Models](https://mlsanity.com/paper/2502.15910/) (2025)
- [Exploring the Personality Traits of LLMs through Latent Features Steering](https://mlsanity.com/paper/2410.10863/) (2024)
- [Model-agnostic Adversarial Attack and Defense for Vision-Language-Action Models](https://mlsanity.com/paper/2510.13237/) (2025)
- [Flattery in Motion: Benchmarking and Analyzing Sycophancy in Video-LLMs](https://mlsanity.com/paper/2506.07180/) (2025)
- [When Modalities Conflict: How Unimodal Reasoning Uncertainty Governs Preference Dynamics in MLLMs](https://mlsanity.com/paper/2511.02243/) (2025)
- [CLIPErase: Efficient Unlearning of Visual-Textual Associations in CLIP](https://mlsanity.com/paper/2410.23330/) (2024)
- [HiPhO: How Far Are (M)LLMs from Humans in the Latest High School Physics Olympiad Benchmark?](https://mlsanity.com/paper/2509.07894/) (2025)
- [Communication Efficient and Provable Federated Unlearning](https://mlsanity.com/paper/2401.11018/) (2024)
- [Model Unlearning via Sparse Autoencoder Subspace Guided Projections](https://mlsanity.com/paper/2505.24428/) (2025)
- [Automating Steering for Safe Multimodal Large Language Models](https://mlsanity.com/paper/2507.13255/) (2025)
- [Goal-oriented Backdoor Attack against Vision-Language-Action Models via Physical Objects](https://mlsanity.com/paper/2510.09269/) (2025)
- [Adaptive Multi-Subspace Representation Steering for Attribute Alignment in Large Language Models](https://mlsanity.com/paper/2508.10599/) (2025)
- [A Hopfieldian View-based Interpretation for Chain-of-Thought Reasoning](https://mlsanity.com/paper/2406.12255/) (2024)
- [Learning to Steer: Input-dependent Steering for Multimodal LLMs](https://mlsanity.com/paper/2508.12815/) (2025)
- [Towards User-level Private Reinforcement Learning with Human Feedback](https://mlsanity.com/paper/2502.17515/) (2025)
- [Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills](https://mlsanity.com/paper/2506.12963/) (2025)
- [Scaling Physical Reasoning with the PHYSICS Dataset](https://mlsanity.com/paper/2506.00022/) (2025)
- [Understanding and Mitigating Cross-lingual Privacy Leakage via Language-specific and Universal Privacy Neurons](https://mlsanity.com/paper/2506.00759/) (2025)

## Built on by (most cited first, 4 of 4)

- [Controlling Repetition in Protein Language Models](https://mlsanity.com/paper/2602.00782/) (2026)
- [Bayesian Gated Non-Negative Contrastive Learning](https://mlsanity.com/paper/2605.28441/) (2026)
- [Multi-Adapter Representation Interventions via Energy Calibration](https://mlsanity.com/paper/2605.28722/) (2026)
- [CUSE: Fine-Grained Multimodal Unlearning with Counterfactual Guidance and On-Policy Self-Distillation](https://mlsanity.com/paper/2610.14763/) (2026)

## Abstract (arXiv)

Machine unlearning aims to erase requested data from trained models without full retraining. For Reasoning Multimodal Large Language Models (RMLLMs), this is uniquely challenging: intermediate chain-of-thought steps can still leak sensitive information even when final answers are forgotten, and overly aggressive interventions easily damage general reasoning ability. Yet no benchmark jointly evaluates how well unlearning methods suppress reasoning-level leakage while preserving reasoning competence. We address this gap with RMLLMU-Bench, the first benchmark for RMLLM unlearning that extends standard forgetting metrics with dedicated measures of reasoning leakage and reasoning retention. A systematic evaluation on RMLLMU-Bench reveals that existing unlearning methods for MLLMs and Large (Language) Reasoning Models (LRMs) either leave substantial leakage in the reasoning process or severely degrade reasoning performance. To address these gaps, we propose R-MUSE (Reasoning-preserving MLLM Unlearning via Subspace guidance and Adaptive Steering), a training-free and inference-time intervention framework that steers internal representations to forget both answers and reasoning traces while explicitly preserving general reasoning. Experiments on RMLLMU-Bench demonstrate that R-MUSE achieves a substantially better balance between effective forgetting and reasoning retention.