Back to AI News
arxiv

ROSE: Risk-Aware Orthogonal Subspace Navigation for Lifelong Knowledge Editing in Multimodal Large Language Models

Researchers propose ROSE, a framework for lifelong knowledge editing in multimodal large language models that restricts parameter updates to the orthogonal complement of previously edited subspaces to reduce interference. ROSE adds a dynamically computed risk mask to protect factual parameter subspaces important for model stability, along with a multi-granularity inference strategy using prototype-anchored semantic gating and perturbation-driven subspace fusion. Experiments on two multimodal benchmarks show the method is competitive with state-of-the-art approaches across five metrics and remains stable over 1,000 sequential updates. Code is publicly released.

AI Brief

It offers a concrete method for sequential multimodal knowledge editing with a stability-focused design and released code, useful to researchers working on model updates without forgetting.

arXiv:2611.09770v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) require continuous factual updates to stay current, yet lifelong knowledge editing remains a daunting challenge. The primary obstacle is the severe Locality Erosion and Catastrophic Forgetting, where sequential edits inevitably interfere with pre-trained capabilities and prior updates. In this paper, we propose \textbf{ROSE}, a \textbf{R}isk-aware \textbf{O}rthogonal \textbf{S}ubspace navigation framework for lifelong knowledge \textbf{E}diting. ROSE addresses forgetting by aiming to restrict parameter updates to the orthogonal complement of previously edited subspaces. Crucially, it introduces a risk-guided mechanism that utilizes a dynamically computed risk mask to safeguard specific factual parameter subspaces essential for model stability. During inference, ROSE employs a multi-granularity knowledge integration strategy, featuring prototype-anchored semantic gating and perturbation-driven subspace fusion to precisely activate edited knowledge for relevant queries while limiting interference on unrelated inputs. Comprehensive experiments on two prominent multimodal benchmarks show that ROSE is competitive with state-of-the-art approaches across five key metrics, while maintaining outstanding stability over $1,000$ sequential updates. Our code is available at https://github.com/sunlitsong/ROSE.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!