Options
PROTOTEX: Explaining Model Decisions with Prototype Tensors
Journal
Proceedings of the Annual Meeting of the Association for Computational Linguistics
ISSN
0736587X
Date Issued
2022-01-01
Author(s)
Das, Anubrata
Gupta, Chitrank
Kovatchev, Venelin
Lease, Matthew
Li, Junyi Jessy
Abstract
We present PROTOTEX, a novel white-box NLP classification architecture based on prototype networks (Li et al., 2018). PROTOTEX faithfully explains model decisions based on prototype tensors that encode latent clusters of training examples. At inference time, classification decisions are based on the distances between the input text and the prototype tensors, explained via the training examples most similar to the most influential prototypes. We also describe a novel interleaved training algorithm that effectively handles classes characterized by the absence of indicative features. On a propaganda detection task, PROTOTEX accuracy matches BART-large and exceeds BERT-large with the added benefit of providing faithful explanations. A user study also shows that prototype-based explanations help non-experts to better recognize propaganda in online news.
Volume
1