发帖
雷达卡
Generative Causal Explanations for Graph Neural Networks
Wanyu Lin 1 2 Hao Lan 2 Baochun Li 2
Abstract Despite their practical success, most GNNs are deployed as
This paper presents Gem, a model-agnostic ap- black boxes, lacking explicit declarative knowledge repre-
proach for providing interpretable explanations sentations. Therefore, they have difficulty in generating the
for any GNNs on various graph learning tasks. ...
京公网安备 11010802022788号
