We argue that we need to evaluate model interpretability methods 'in the wild', i.e., in situations where professionals make critical decisions, and models can potentially assist them. We present an in-the-wild evaluation of token attribution based on Deep Taylor Decomposition, with professional journalists performing reliability assessments. We find that using this method in conjunction with RoBERTa-Large, fine-tuned on the Gossip Corpus, led to faster and better human decision-making, as well as a more critical attitude toward news sources among the journalists. We present a comparison of human and model rationales, as well as a qualitative analysis of the journalists' experiences with machine-in-the-loop decision making.
Cite
@article{arxiv.2206.02661,
title = {Evaluating Deep Taylor Decomposition for Reliability Assessment in the Wild},
author = {Stephanie Brandl and Daniel Hershcovich and Anders Søgaard},
journal= {arXiv preprint arXiv:2206.02661},
year = {2022}
}