This paper presents Hierarchical Network Dissection, a general pipeline to interpret the internal representation of face-centric inference models. Using a probabilistic formulation, Hierarchical Network Dissection pairs units of the model with concepts in our "Face Dictionary" (a collection of facial concepts with corresponding sample images). Our pipeline is inspired by Network Dissection, a popular interpretability model for object-centric and scene-centric models. However, our formulation allows to deal with two important challenges of face-centric models that Network Dissection cannot address: (1) spacial overlap of concepts: there are different facial concepts that simultaneously occur in the same region of the image, like "nose" (facial part) and "pointy nose" (facial attribute); and (2) global concepts: there are units with affinity to concepts that do not refer to specific locations of the face (e.g. apparent age). To validate the effectiveness of our unit-concept pairing formulation, we first conduct controlled experiments on biased data. These experiments illustrate how Hierarchical Network Dissection can be used to discover bias in the training data. Then, we dissect different face-centric inference models trained on widely-used facial datasets. The results show models trained for different tasks have different internal representations. Furthermore, the interpretability results reveal some biases in the training data and some interesting characteristics of the face-centric inference tasks.

翻译：本文展示了等级网络剖析, 这是一种解释以表心推导模型内部代表性的一般管道。 使用概率配方, 等级网络剖析将模型的单元与“ 面部字典” 中的概念相配( 收集面部概念, 并附相应的样本图像 ) 。 我们的管道受网络分解的启发, 即: 对象中心模型和场景中心模型的流行解释模型 。 然而, 我们的配方可以应对两个重要挑战, 网络分解无法解决 : (1) 概念的平和重叠: 在图像的同一区域同时出现不同的面部概念, 比如“ 鼻部”( 缩略) 和“ 点鼻” ( 缩略属性) ; (2) 全球概念: 网络分解是来自网络分解的, 即: 网络分解模式与对象中心对象中心模型不相近( 明显年龄 ) 。 为了验证我们单位对立的配方模型的有效性, 我们首先对偏向性数据进行受控的实验。 这些实验表明, 纵向网络剖面剖分解可如何在培训数据中同时发现数据中的偏向性, 。 之后, 我们经过培训的面面部位分析后, 将一些分析的外观分析结果, 显示结果 。