BriefGPT.xyz
Mar, 2022
在多模态BERT中发现结构知识
Finding Structural Knowledge in Multimodal-BERT
HTML
PDF
Victor Milewski, Miryam de Lhoneux, Marie-Francine Moens
TL;DR
本文探讨了多模态 BERT 模型嵌入学习到的知识，并通过依存句法分析语言数据和对象数据之间的依赖关系，构建了显式的可视化场景树，最终实验证明了多模态 BERT 模型不能编码这些场景树。
Abstract
In this work, we investigate the knowledge learned in the embeddings of
multimodal-bert
models. More specifically, we probe their capabilities of storing the
grammatical structure
of linguistic data and the struc
→