Skip to content

Latest commit

 

History

History
39 lines (27 loc) · 1.33 KB

README.md

File metadata and controls

39 lines (27 loc) · 1.33 KB
Raw

ProtoTEx: Explaining Model Decisions with Prototype Tensors

This repository contains example implementations of Prototype tensors for providing case-based reasoning.

Find a copy of the paper here: https://utexas.app.box.com/v/das-acl-2022

NOTE: This code is written to reproduce the experimental results reportedin the paper. If you run into any issues, please contact me.

Quick Start

To run ProtoTEx on the propganda detection task, first we need the following:

mkdir Logs
mkdir Models
pip install -r requirements.txt

If pytorch installation fails with the above command, please follow the instructions here and then rerun pip install.

For a quick run of the model, run the notebook train-eval-propaganda.ipynb.

Citation

If you use this resource or ideas presented here, please consider citing our paper:

BibTeX

@inproceedings{das-acl22,
  author = {Anubrata Das and Chitrank Gupta and Venelin Kovatchev and Matthew Lease and Junyi Jessy Li},
  title = {{{\sc ProtoTEx}: Explaining Model Decisions with Prototype Tensors}},
  booktitle = {{Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (ACL)}},
  year = {2022},
  url = {https://utexas.box.com/v/das-acl-2022},
  source = {https://github.com/anubrata/ProtoTEx/},
  note = {12 pages.}
}