δ-CLUE: Diverse Sets of Explanations for Uncertainty Estimates

arxiv(2021)

引用 0|浏览2
暂无评分
摘要
To interpret uncertainty estimates from differentiable probabilistic models, recent work has proposed generating Counterfactual Latent Uncertainty Explanations (CLUEs). However, for a single input, such approaches could output a variety of explanations due to the lack of constraints placed on the explanation. Here we augment the original CLUE approach, to provide what we call δ-CLUE. CLUE indicates one way to change an input, while remaining on the data manifold, such that the model becomes more confident about its prediction. We instead return a set of plausible CLUEs: multiple, diverse inputs that are within a δ ball of the original input in latent space, all yielding confident predictions.
更多
查看译文
关键词
uncertainty estimates,explanations
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要