SGPVT: Self-Generated Proximal Visual Tokens for Mitigating Proximal Collateral Damage in MLLM Unlearning

ACL ARR 2026 January Submission1268 Authors

Keywords: Machine unlearning, Multimodal Large Language Model

Abstract: Machine unlearning in multimodal large language models (MLLMs) aims to remove specific concepts while preserving overall utility. However, existing approaches focus primarily on general utility metrics, overlooking the preservation of semantically related concepts. We present the first systematic analysis of this proximal collateral damage, revealing that forgetting vulnerability correlates strongly with visual embedding similarity in a smooth gradient across the semantic space. Based on this insight, we propose a novel unlearning framework that introduces Self-Generated Proximal Visual Tokens (SGPVTs), which are synthetically perturbed visual representations around the target concept. Our method employs an adaptive cosine-band curriculum with a dual-stream objective: forgetting the target via gradient ascent while distilling knowledge from a frozen teacher model into proximal tokens to prevent degradation. Extensive experiments demonstrate that our approach significantly outperforms existing methods in preserving semantically related concepts while achieving effective target unlearning, eliminating the need for manual retention set curation. Our source code will be released in the near future.

Paper Type: Long

Research Area: Multimodality and Language Grounding to Vision, Robotics and Beyond

Research Area Keywords: Machine unlearning, Multimodal Large Language Model

Contribution Types: NLP engineering experiment, Publicly available software and/or pre-trained models, Data resources

Languages Studied: English

Submission Number: 1268