TabPert: An Effective Platform for Tabular Perturbation

TL;DR

TabPert is a platform that generates counterfactual tabular data to evaluate and analyze the reasoning capabilities of Natural Language Inference models, highlighting their strengths and weaknesses.

Contribution

It introduces a comprehensive tool for creating and analyzing counterfactual tables and hypotheses, aiding in systematic model evaluation.

Findings

Enables generation of challenging counterfactual data

Provides detailed metadata for analysis

Facilitates identification of model shortcomings

Abstract

To truly grasp reasoning ability, a Natural Language Inference model should be evaluated on counterfactual data. TabPert facilitates this by assisting in the generation of such counterfactual data for assessing model tabular reasoning issues. TabPert allows a user to update a table, change its associated hypotheses, change their labels, and highlight rows that are important for hypothesis classification. TabPert also captures information about the techniques used to automatically produce the table, as well as the strategies employed to generate the challenging hypotheses. These counterfactual tables and hypotheses, as well as the metadata, can then be used to explore an existing model's shortcomings methodically and quantitatively.