作者
Suraj Srinivas, François Fleuret
发表日期
2019/5/2
期刊
Neural Information Processing Systems (NeurIPS)
简介
We introduce a new tool for interpreting neural nets, namely full-gradients, which decomposes the neural net response into input sensitivity and per-neuron sensitivity components. This is the first proposed representation which satisfies two key properties: completeness and weak dependence, which provably cannot be satisfied by any saliency map-based interpretability method. Using full-gradients, we also propose an approximate saliency map representation for convolutional nets dubbed FullGrad, obtained by aggregating the full-gradient components.
学术搜索中的文章
S Srinivas, F Fleuret - Advances in neural information processing systems, 2019
S Suraj, F Francois - arXiv preprint arxiv:1905.00780, 2019