The attention mechanism plays an important role in the machine reading comprehension (MRC) model. Here, we describe a pipeline for building an MRC model with a pretrained language model and visualizing the effect of each attention zone in different layers, which can indicate the explainability of the model. With the presented protocol and accompanying code, researchers can easily visualize the relevance of each attention zone in the MRC model. This approach can be generalized to other pretrained language models.
@article{arxiv.2410.20652,
title = {Visualizing attention zones in machine reading comprehension models},
author = {Yiming Cui and Wei-Nan Zhang and Ting Liu},
journal= {arXiv preprint arXiv:2410.20652},
year = {2024}
}