arXiv 2509.05714

Towards Meta-Cognitive Knowledge Editing for Multimodal LLMs

By Zhaoyu Fan, Kaihang Pan, et al.

Published 2025-09-06

Discussion

Read the public discussion and references gathered around this paper.

Knowledge editing enables multimodal large language models (MLLMs) to efficiently update outdated or incorrect information.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…

View the original paper on arXiv