Describe-and-Dissect: Interpreting Neurons in Vision Networks with Language Models

arXiv:2403.13771v1 Announce Type: cross
Abstract: In this paper, we propose Describe-and-Dissect (DnD), a novel method to describe the roles of hidden neurons in vision networks. DnD utilizes recent advancements in multimodal deep learning to produce complex natural language descriptions, without the need for labeled training data or a predefined set of concepts to choose from. Additionally, DnD is training-free, meaning we don't train any new models and can easily leverage more capable general purpose models in the future. We have …

abstract arxiv cs.cv cs.lg data deep learning hidden language language models multimodal multimodal deep learning natural natural language networks neurons novel paper roles set training training data type vision