论文标题
与神经网络进行比较的反面表示相似性
Deconfounded Representation Similarity for Comparison of Neural Networks
论文作者
论文摘要
相似性指标（例如代表性相似性分析（RSA）和中心内核比对（CKA））已用于比较神经网络之间的层表示。但是，这些指标被输入空间中数据项的种群结构混淆，从而导致甚至完全随机的神经网络和转移学习中的域关系不一致的高度相似性。我们引入了一个简单且通常适用的修复程序，以通过协变量调整回归为混杂因素进行调整，该回归保留了原始相似性度量的直观不变属性。我们表明，反相似度指标的反面增加了检测语义上相似的神经网络的分辨率。此外，在现实世界中，变形可以提高表示与转移学习中域相似性的表示相似性的一致性，并提高了与分布式准确性的相关性。
Similarity metrics such as representational similarity analysis (RSA) and centered kernel alignment (CKA) have been used to compare layer-wise representations between neural networks. However, these metrics are confounded by the population structure of data items in the input space, leading to spuriously high similarity for even completely random neural networks and inconsistent domain relations in transfer learning. We introduce a simple and generally applicable fix to adjust for the confounder with covariate adjustment regression, which retains the intuitive invariance properties of the original similarity measures. We show that deconfounding the similarity metrics increases the resolution of detecting semantically similar neural networks. Moreover, in real-world applications, deconfounding improves the consistency of representation similarities with domain similarities in transfer learning, and increases correlation with out-of-distribution accuracy.
