Explaining Latent Representations with a Corpus of Examples

• DOI: 10.48550/arxiv.2110.15355
• Publication Date: 2021-01-01
• Authors: Jonathan Crabbé; Zhaozhi Qian; Fergus Imrie; Mihaela van der S...
• Keywords: Representation; Simplex; Text corpus

Modern machine learning models are complicated. Most of them rely on convoluted latent representations their input to issue a prediction. To achieve greater transparency than black-box that connects inputs predictions, it is necessary gain deeper understanding these representations. aim, we propose SimplEx: user-centred method provides example-based explanations with reference freely selected set examples, called the corpus. SimplEx uses corpus improve user's space post-hoc answering two questions: (1) Which examples explain prediction issued for given test example? (2) What features relevant model relate an answer by reconstructing representation as mixture Further, novel approach, Integrated Jacobian, allows make explicit contribution each feature in mixture. Through experiments tasks ranging from mortality image classification, demonstrate decompositions robust and accurate. With illustrative use cases medicine, show empowers user highlighting patterns Moreover, how freedom choosing have personalized terms meaningful them.