aiqbench analysis
This paper critically examines flawed evaluation protocols in machine unlearning for multimodal large language models, exposing how current benchmarks conflate genuine forgetting with instance-wise overfitting and rely on synthetic data. The authors introduce GEMMU, a new benchmark utilizing real-world knowledge to reveal significant generalization gaps that undermine previous conclusions about unlearning efficacy. These findings are essential for researchers aiming to develop robust, trustworthy multimodal systems where data privacy and selective memory retention are paramount.
Are We Really Measuring Forgetting? Pitfalls in Evaluating Machine Unlearning for Multimodal LLMs
arXiv:2610.19613v1 Announce Type: new Abstract: As machine unlearning for multimodal large language models (MLLMs) rapidly advances, rigorous evaluation of its effectiveness becomes increasingly important. However, existing benchmarks largely inherit evaluation protocols from text-only unlearning and overlook the unique challenges of unlearning in multimodal settings. In this work, we identify three major pitfalls that may lead to misleading conclusions: the lack of heldout evaluation to distinguish genuine concept-wise forgetting from instance-wise overfitting, the reliance on synthetic, fine-tuned knowledge rather than real-world knowledge acquired during pre-training, and unfair method comparisons caused by applying the same training settings to different unlearning methods. To address these challenges, we introduce GEMMU (Generalizable Evaluation for Multimodal Machine Unlearning), a benchmark built on real-world knowledge and a principled evaluation protocol. Re-evaluating representative unlearning methods under GEMMU reveals substantial generalization gaps and challenges conclusions drawn from previous benchmarks. Our findings highlight the importance of rigorous evaluation for advancing multimodal machine unlearning and provide guidelines for future benchmark design.
Title:Are We Really Measuring Forgetting? Pitfalls in Evaluating Machine Unlearning for Multimodal LLMs
Abstract:As machine unlearning for multimodal large language models (MLLMs) rapidly advances, rigorous evaluation of its effectiveness becomes increasingly important. However, existing benchmarks largely inherit evaluation protocols from text-only unlearning and overlook the unique challenges of unlearning in multimodal settings. In this work, we identify three major pitfalls that may lead to misleading conclusions: the lack of heldout evaluation to distinguish genuine concept-wise forgetting from instance-wise overfitting, the reliance on synthetic, fine-tuned knowledge rather than real-world knowledge acquired during pre-training, and unfair method comparisons caused by applying the same training settings to different unlearning methods. To address these challenges, we introduce GEMMU (Generalizable Evaluation for Multimodal Machine Unlearning), a benchmark built on real-world knowledge and a principled evaluation protocol. Re-evaluating representative unlearning methods under GEMMU reveals substantial generalization gaps and challenges conclusions drawn from previous benchmarks. Our findings highlight the importance of rigorous evaluation for advancing multimodal machine unlearning and provide guidelines for future benchmark design.
Computation and Language (cs.CL)
arXiv:2610.19613 [cs.CL]
(or arXiv:2610.19613v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2610.19613 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 263 words
Related coverage
Reflections on unlearning and inoculation
LessWrong · Sep 20, 2026
Presentation: Running AI at the Edge: Running Real Workloads Directly in the Browser
InfoQ · Aug 31, 2026
Runway’s WorldPrompt and the Engineering of Real-Time Worlds
Latent Space · Sep 25, 2026
Do LLMs Really Forget? Hidden-State Leakage in Model Unlearning and How to Fix it
arXiv · Sep 30, 2026