arXiv 2509.05714
Towards Meta-Cognitive Knowledge Editing for Multimodal LLMs
By Zhaoyu Fan, Kaihang Pan, et al.
Published 2025-09-06
Mindmap
Browse the paper's core ideas, clusters, and relationships in a structured outline.
Knowledge editing enables multimodal large language models (MLLMs) to efficiently update outdated or incorrect information. However, existing benchmarks primarily emphasize cognitive-level modifications while lacking a focus on deeper meta-cognitive processes. To bridge this gap, we introduce CogEdit, a novel benchmark designed to evaluate MLLMs' meta-cognitive knowledge editing abilities across three levels: (1) Co…