Reasoning-Aware Cross-Modal Association Unlearning in Multimodal Large Reasoning Models

Xinghao Li; Chenchen Tan; Youyang Qu; Shujie Cui; Xingliang Yuan; Cunjian Chen; Chi Liu; Longxiang Gao

cs.CVNov 4, 2026

Reasoning-Aware Cross-Modal Association Unlearning in Multimodal Large Reasoning Models

Authors: Xinghao Li, Chenchen Tan, Youyang Qu, Shujie Cui, Xingliang Yuan, Cunjian Chen, Chi Liu, Longxiang Gao

Organizations: Faculty of Information Technology, Monash University, Australia · Qilu University of Technology (Shandong Academy of Sciences), China · Shandong Computer Science Center (National Supercomputer Center in Jinan), China · School of Computing and Information Systems, The University of Melbourne, Australia ·

Abstract

Multimodal large reasoning models (MLRMs) integrate visual inputs with textual and parametric knowledge, allowing sensitive information to be accessed through cross-modal associations during reasoning. This raises privacy concerns when a visual input should no longer provide access to such information. Machine unlearning can remove this influence without retraining the model from scratch. However, multimodal unlearning methods mainly target samples, concepts, or knowledge content, without separating the knowledge from the cross-modal association through which it is accessed. In reasoning models, target information may disappear from the final answer while remaining in intermediate reasoning, or be replaced by incorrect substitute knowledge. We therefore study cross-modal association unlearning, where the forget target is a designated visual-to-knowledge association: the model should suppress target-bearing behavior under the designated visual context while preserving access to the same knowledge in permitted contexts. We propose Visual-Keyed Counterfactual Rebinding (VKCR), a parameter-efficient framework for cross-modal association editing. VKCR first localizes the edit to a compact association-conditioned parameter subspace. It then reduces support for the designated target at task outputs and counterfactually rebinds the affected reasoning behavior by discouraging target-bearing or unsupported continuations while favoring valid non-sensitive alternatives. We further introduce CLEAR-R and MLLMU-R to evaluate target leakage, hallucinated reassociation, reasoning preservation, and modality-specific utility. Experiments show that VKCR substantially reduces image-conditioned target leakage and hallucinated reassociation, including across heterogeneous task formulations, while preserving non-sensitive visual reasoning and permitted text-only knowledge.

Figures & tables

Figure 1: Cross-modal association unlearning. The goal is to suppress a designated image-conditioned association to target knowledge while preserving permissible text-only knowledge. Existing methods may leave reasoning leakage or induce hallucinated reassociation, whereas VKCR selectively de-binds the sensitive visual association while retaining non-sensitive reasoning and the underlying knowledge.

Method	CLEAR-R					MLLMU-R
Method	Qwen2.5-VL-3B-Instruct					LLaVA-1.5-7B-HF
	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$
Vanilla	30.16	50.28	36.47	92.09	27.79	73.11	98.94	22.48	97.62	68.86
Grad. Diff. ( Liu et al., 2022 )	26.91	20.90	57.34	93.07	26.74	70.77	66.61	56.09	84.92	64.98
KL-Min ( Nguyen et al., 2020 )	28.23	36.72	41.49	95.52	27.52	56.95	78.10	54.04	74.97	49.53
NPO ( Zhang et al., 2024 )	27.89	36.16	52.55	90.18	27.90	54.51	86.94	38.39	79.17	51.38

Table 1: Main results under the 5% forgetting setting in CLEAR-R and MLLMU-R on generation task. F-RL/R-RL : forget/retain ROUGE-L; F-RSL : reasoning-sensitive leakage; F-RSH : reasoning hallucination; F-RSP : non-sensitive reasoning preservation.

\downarrow

: lower is better;

\uparrow

: higher is better.

Figure 2: Selective association de-binding under the 5% forgetting setting. Img-K measured by ROUGE-L on multimodal question answering for the forget set. Txt-K measures access to the same target knowledge through pure-text questions, also using ROUGE-L. Vis-Ret measures preservation of non-sensitive visual information, i.e., how much of the image content unrelated to the forgotten private information can still be correctly described. Ret. denotes retain-set utility, measured by retain-set ROUGE-L.

Figure 3: Cross-task unlearning effectiveness on MLLMU-R under the 5% forgetting setting. We compare final-answer forgetting (x-axis) with reasoning-sensitive leakage (y-axis) across classification, generation, and cloze tasks. Lower values are desirable on both axes.

MLLMU-R: perturbation recovery
Task	Clean	Query	Vis. Noise	Joint
Classification	69.10	68.12	66.72	64.90
Generation	35.28	34.12	32.83	30.64
Cloze	32.80	31.28	29.87	27.98
CLEAR-R: on-key vs. off-key control
Visual condition	BASE	VKCR	$\bm{\Delta}$	$\bm{\rho}$

Table 2: Cross-modal perturbation and visual-key control under the 5% forgetting setting. All values are F-RSL (%, lower is better). MLLMU-R tests query and visual perturbations, while CLEAR-R compares on-key and off-key visual conditions.

Variant	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$
w/o Association-Conditioned	18.53	27.72	31.51	88.72	41.87
w/o On-Policy Refinement	41.18	63.29	27.29	95.31	61.23
w/o Reassociation Control	27.50	35.50	59.00	91.50	55.47
w/o Behavioral Protection	21.86	28.38	22.19	81.76	27.94
Full VKCR	31.25	35.28	21.25	94.81	59.37

Table 3: Ablation study of VKCR on MLLMU-R under the 5% forgetting setting. F-RL measures final-answer forgetting, F-RSL reasoning-sensitive leakage, F-RSH hallucinated reassociation, F-RSP non-sensitive reasoning preservation, and R-RL retain utility. Lower is better for F-RL/F-RSL/F-RSH; higher is better for F-RSP/R-RL.

Figure 4: Generation task unlearning performance on the R1-Onevision-7B-RL reasoning backbone under the 5% forgetting setting. For CLEAR-R and MLLMU-R, the forgetting panels plot F-RSL against F-RSH, where lower values indicate cleaner suppression of the target association, while the preservation panels plot F-RSP against generation-task R-RSP, where higher values indicate better reasoning coherence preservation.

Appendix figures & tables16 assets

Supplementary material from the paper’s appendix.

Appendix

Method	Main Hyperparameters
Grad. Diff.	epochs $=2$ , lr $=3\times 10^{-5}$ , batch size $=4$
KL-Min	epochs $=2$ , lr $=3\times 10^{-5}$ , batch size $=4$
NPO	epochs $=2$ , lr $=3\times 10^{-5}$ , $\beta=0.4$ , batch size $=4$
MANU	pruning ratio $=10\%$ , dual-modality pruning, $\tau=0.1$
R2MU	lr $=7.5\times 10^{-5}$ , $\alpha=1.4$ , steering coefficient $=6.5$
MMUnlearner	lr $=1\times 10^{-5}$ , epochs $=1$ , forget coefficient $=1.0$

Appendix

Table 4: Main hyperparameters on CLEAR-R.

Method	Main Hyperparameters
Grad. Diff.	epochs $=2$ , lr $=2\times 10^{-5}$ , batch size $=4$
KL-Min	epochs $=2$ , lr $=3\times 10^{-5}$ , batch size $=4$
NPO	epochs $=2$ , lr $=3\times 10^{-5}$ , $\beta=0.6$ , batch size $=4$
MANU	pruning ratio $=10\%$ , dual-modality pruning, $\tau=0.1$
R2MU	lr $=7.5\times 10^{-5}$ , $100$ steps, $\alpha=2.0$ , steering coefficient $=6.5$
MMUnlearner	lr $=1\times 10^{-5}$ , epochs $=1$ , forget coefficient $=1.5$

Appendix

Table 5: Main hyperparameters on MLLMU-R.

Metric	Split	Primary Failure Mode	Direction
F-RSL	Forget	Target-sensitive reasoning leakage	$\downarrow$
F-RSH	Forget	Unsupported/hallucinated reasoning	$\downarrow$
F-RSP	Forget	Non-sensitive reasoning preservation	$\uparrow$
R-RSP	Retain	Retain-side reasoning preservation	$\uparrow$

Appendix

Table 6: Summary of the reasoning-aware metrics.

Dataset	Backbone	Method	GPU-hours	Updated Params	% of Model	Update Type
CLEAR-R	Qwen2.5-VL-3B- Instruct	GA-Diff	2.098	14,966,784	$\sim$ 0.40%	LoRA language params
		KL-Min	2.912	14,966,784	$\sim$ 0.40%	LoRA language params
		NPO	2.062	14,966,784	$\sim$ 0.40%	LoRA language params
		MANU	1.157	0	0%	Activation/channel pruning
		R 2 MU	1.595	940,032	$\sim$ 0.025%	Selected LoRA MLP params
		MMUnlearner	2.466	14,966,784	$\sim$ 0.40%	LoRA language params

Appendix

Table 7: Computational cost and parameter updates under the 5% forgetting setting. We report GPU-hours, the number and proportion of updated parameters, and the corresponding update type.

Method	CLEAR-R (10%)					MLLMU-R (10%)
Method	Qwen2.5-VL-3B-Instruct					LLaVA-1.5-7B-HF
	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$
Vanilla	29.49	49.01	48.39	97.93	27.93	72.61	91.28	16.45	96.97	68.65
Grad. Diff.	23.78	18.03	59.21	90.82	25.05	44.40	63.09	64.37	75.70	49.20
KL-Min	26.13	27.89	56.31	93.83	26.56	61.95	71.18	65.87	87.72	62.11
NPO	27.53	30.70	54.72	95.46	27.70	60.21	80.17	32.06	83.31	56.86

Appendix

Table 8: Main results under the 10% forgetting setting.

\downarrow

: lower is better;

\uparrow

: higher is better.

Method	CLEAR-R (1% Forgetting)					MLLMU-R (15% Forgetting)
Method	Qwen2.5-VL-3B-Instruct					LLaVA-1.5-7B-HF
	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$
Vanilla	29.16	59.38	42.86	94.71	28.14	67.95	90.68	20.82	97.61	68.89
Grad. Diff.	26.19	43.75	47.14	92.94	27.50	21.36	58.21	47.69	36.79	21.85
KL-Min	28.97	53.13	48.57	91.76	27.89	54.84	73.78	44.03	81.72	56.69
NPO	27.32	56.25	44.29	87.88	26.97	61.62	82.75	31.92	95.09	63.12

Appendix

Table 9: Results under dataset-specific additional forgetting settings. CLEAR-R is evaluated at 1% forgetting, while MLLMU-R is evaluated at 15% forgetting.

\downarrow

: lower is better;

\uparrow

: higher is better.

Method	CLEAR-R					MLLMU-R
Method	R1-Onevision-7B-RL					R1-Onevision-7B-RL
	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$	F-RL $\downarrow$	F-RSL $\downarrow$	F-RSH $\downarrow$	F-RSP $\uparrow$	R-RL $\uparrow$
Vanilla	25.79	10.17	73.92	92.91	25.41	64.50	54.27	76.07	89.62	63.79
Grad. Diff.	16.50	5.65	62.77	81.71	14.92	60.10	51.09	78.82	79.62	59.51
KL-Min	12.12	3.95	49.47	41.79	11.66	55.80	38.17	76.39	71.40	57.49
NPO	24.14	9.60	78.94	83.83	24.02	60.40	46.39	78.03	85.77	62.63

Appendix

Table 10: Additional evaluation on R1-Onevision-7B-RL under the 5% forgetting setting. We evaluate the same unlearning methods on a reasoning-oriented multimodal backbone. Lower is better for F-RL, F-RSL, and F-RSH, while higher is better for F-RSP and R-RL.

Figure 5: Retain-side reasoning preservation under different forgetting ratios. R-RSP measures preserved reasoning coherence on retained examples, where higher is better. Circles and squares denote multimodal and text-only settings, respectively.

Method	CoT (key excerpt)	Final answer (key excerpt)	Behavior
Grad. Diff.	“… explicitly identifies the person as Zeynab Nazirova …”	“The person is Zeynab Nazirova , shown standing outdoors wearing an orange and green headscarf …”	Hallucinated identity substitution
KL-Min	“… explicitly identifies the person as Hina Ameen …”	“The person is Hina Ameen , shown standing outdoors in front of a rocky landscape and holding an open book.”	Direct CoT leakage
NPO	“…explicitly identifies the person as Ji-Yeon Soo …”	“The person is Ji-Yeon Soo , shown standing outdoors in front of a rocky landscape and holding an open book.”	Hallucinated identity substitution
MANU	“… explicitly names the person as Zeynab Nazirova …”	“The person is Zeynab Nazirova , shown standing outdoors in front of a vast, rocky landscape …”	Hallucinated identity substitution
R 2 MU	“… explicitly identifies the person as Zeynab Nazirova …”	“The person is Zeynab Nazirova , shown standing outdoors in a rocky landscape and holding a book …”	Hallucinated identity substitution
MMUnlearner	“… explicitly identifies the person as Jambo Mpendulo and states that this person is holding a book titled Jambo: A Tale of the Crumbling City .”	“The person is Jambo Mpendulo …holding a book titled Jambo: A Tale of the Crumbling City .”	Identity + fact hallucination

Appendix

Table 11: Qualitative comparison on unlearned model prediction in CLEAR-R. Existing unlearning methods either retain the target identity in the reasoning trace or replace it with an unsupported identity. In contrast, VKCR removes the target identity while preserving non-sensitive visual evidence such as appearance, objects, and scene context.

Task	Count
SFT QA	8,201
Classification	11,500
Generation	4,608
Cloze	4,610
Total	28,919

Appendix

Table 12: Final MLLMU-R composition after reasoning-level adjudication.

Panel of Table 13 — Table 13: Prompt used to guide GPT-5.4 to generate reasoning annotations for CLEAR-R multimodal examples.

Panel of Table 14 — Table 14: Prompt used to guide GPT-5.4 to generate reasoning annotations for MLLMU-R.

Panel of Table 15 — Table 15: Prompt used to guide GPT-5.4-mini to evaluate implicit target-linked reasoning leakage for RSL.

Panel of Table 16 — Table 16: Prompt used to guide GPT-5.4-mini to evaluate implicit hallucinated or unsupported reasoning for RSH.

Panel of Table 17 — Table 17: Prompt used to guide GPT-5.4-mini to evaluate semantic preservation of retained reasoning steps for F-RSP.

Panel of Table 18 — Table 18: Prompt used to guide GPT-5.4-mini to evaluate retained reasoning support preservation for R-RSP.

MLLM Unlearning

Multimodal large language model (MLLM) unlearning studies how to remove selected knowledge or behaviors while preserving other capabilities.

See all 40 papers

LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection
Aug 12, 2026Xinhao Zhong, Yuxia Qiao, Junhao Li +3VLM Unlearning Multimodal Large Language Models
3Pengcheng Laboratory · 1Harbin Institute of Technology, Shenzhen · 2Tsinghua University
Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks
Jul 8, 2026Nobin Sarwar, Shubhashis Roy Dipta, Zheyuan Liu +1VLM Unlearning MLLM Unlearning
University of Maryland, Baltimore County · University of Notre Dame · UNC Chapel Hill
One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models
Jul 17, 2026Sudharshan Balaji, Yili Ren, Guangjing Wang +2VLM Unlearning VLM Robustness
University of South Florida · USA · University of Massachusetts Lowell

Aug 12, 2026J/K move · Enter open · S save

LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection

Xinhao Zhong, Yuxia Qiao, Junhao Li, Hao Fang+2

3Pengcheng Laboratory · 1Harbin Institute of Technology, Shenzhen · 2Tsinghua University

VLM Unlearning Multimodal Large Language Models Privacy Leakage in Language Models MLLM Unlearning Machine Unlearning

Open paper AskPDF arXiv

Reinforcement-learning (RL) post-training equips multimodal large reasoning models (MLRMs) with exploratory chains of thought (CoT), substantially improving visual reasoning. However, we find that this capability introduces a distinct privacy vulnerability: even when a sensitive fact is successfully unlearned from the final answer, the model may still reproduce it in its reasoning trace. This leakage is substantially more pronounced in natively RL-trained MLRMs than in their non -reasoning base models, revealing a privacy risk that existing unlearning methods are not designed to address. We show that RL-induced exploration leaves sensitive content with a distinctive token-level entropy signature that is largely absent from base models. Based on this observation, we propose LEMUR, a fully training-free, inference-time unlearning framework for natively RL-trained multimodal models. LEMUR uses entropy dynamics as a control signal to identify when sensitive reasoning begins and when sanitization should stop. During this interval, it redirects the reasoning trajectory through entropy-modulated visual-anchor latent injection, replacing committed tokens with sanitized, probability-weighted embeddings re-grounded in the input image. Across diverse MLRMs, LEMUR consistently outperforms existing unlearning met hods in suppressing both reasoning-trace and answer leakage, while better preserving non-sensitive utility and output fluency. These results demonstrate that RL-induced entropy dynamics provide a distinctive signal for privacy leakage and that exploiting this signal enables effective training-free unlearning for reasoning-capable multimodal models.

Abstract

MLLM Unlearning

Explore similar work

LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection

Multimodal Unlearning Across Vision, Language, Video, and Audio: Survey of Methods, Datasets, and Benchmarks

One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models