OFFSIDE: Benchmarking Unlearning Misinformation in Multimodal Large Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zheng, Hao, Pang, Zirui, li, Ling, Deng, Zhijie, Pu, Yuhan, Zhu, Zhaowei, Xia, Xiaobo, Wei, Jiaheng
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909980777512960
author Zheng, Hao
Pang, Zirui
li, Ling
Deng, Zhijie
Pu, Yuhan
Zhu, Zhaowei
Xia, Xiaobo
Wei, Jiaheng
author_facet Zheng, Hao
Pang, Zirui
li, Ling
Deng, Zhijie
Pu, Yuhan
Zhu, Zhaowei
Xia, Xiaobo
Wei, Jiaheng
contents Advances in Multimodal Large Language Models (MLLMs) intensify concerns about data privacy, making Machine Unlearning (MU), the selective removal of learned information, a critical necessity. However, existing MU benchmarks for MLLMs are limited by a lack of image diversity, potential inaccuracies, and insufficient evaluation scenarios, which fail to capture the complexity of real-world applications. To facilitate the development of MLLMs unlearning and alleviate the aforementioned limitations, we introduce OFFSIDE, a novel benchmark for evaluating misinformation unlearning in MLLMs based on football transfer rumors. This manually curated dataset contains 15.68K records for 80 players, providing a comprehensive framework with four test sets to assess forgetting efficacy, generalization, utility, and robustness. OFFSIDE supports advanced settings like selective unlearning and corrective relearning, and crucially, unimodal unlearning (forgetting only text data). Our extensive evaluation of multiple baselines reveals key findings: (1) Unimodal methods (erasing text-based knowledge) fail on multimodal rumors; (2) Unlearning efficacy is largely driven by catastrophic forgetting; (3) All methods struggle with "visual rumors" (rumors appear in the image); (4) The unlearned rumors can be easily recovered and (5) All methods are vulnerable to prompt attacks. These results expose significant vulnerabilities in current approaches, highlighting the need for more robust multimodal unlearning solutions. The code is available at https://github.com/zh121800/OFFSIDE
format Preprint
id arxiv_https___arxiv_org_abs_2510_22535
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle OFFSIDE: Benchmarking Unlearning Misinformation in Multimodal Large Language Models
Zheng, Hao
Pang, Zirui
li, Ling
Deng, Zhijie
Pu, Yuhan
Zhu, Zhaowei
Xia, Xiaobo
Wei, Jiaheng
Artificial Intelligence
Computation and Language
Advances in Multimodal Large Language Models (MLLMs) intensify concerns about data privacy, making Machine Unlearning (MU), the selective removal of learned information, a critical necessity. However, existing MU benchmarks for MLLMs are limited by a lack of image diversity, potential inaccuracies, and insufficient evaluation scenarios, which fail to capture the complexity of real-world applications. To facilitate the development of MLLMs unlearning and alleviate the aforementioned limitations, we introduce OFFSIDE, a novel benchmark for evaluating misinformation unlearning in MLLMs based on football transfer rumors. This manually curated dataset contains 15.68K records for 80 players, providing a comprehensive framework with four test sets to assess forgetting efficacy, generalization, utility, and robustness. OFFSIDE supports advanced settings like selective unlearning and corrective relearning, and crucially, unimodal unlearning (forgetting only text data). Our extensive evaluation of multiple baselines reveals key findings: (1) Unimodal methods (erasing text-based knowledge) fail on multimodal rumors; (2) Unlearning efficacy is largely driven by catastrophic forgetting; (3) All methods struggle with "visual rumors" (rumors appear in the image); (4) The unlearned rumors can be easily recovered and (5) All methods are vulnerable to prompt attacks. These results expose significant vulnerabilities in current approaches, highlighting the need for more robust multimodal unlearning solutions. The code is available at https://github.com/zh121800/OFFSIDE
title OFFSIDE: Benchmarking Unlearning Misinformation in Multimodal Large Language Models
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2510.22535