# Distribution Matching for Rationalization

Source: [Distribution Matching for Rationalization](<https://www.sota2.com/research/paper/68fa6a0f1e209131cea25e22>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:56:48.130Z

Authors: Yongfeng Huang, Yujun Chen, Yulun Du, Zhilin Yang

Published: 2021-06-01T08:49:32.000Z

## Abstract

The task of rationalization aims to extract pieces of input text as rationales to justify neural network predictions on text classification tasks. By definition, rationales represent key text pieces used for prediction and thus should have similar classification feature distribution compared to the original input text. However, previous methods mainly focused on maximizing the mutual information between rationales and labels while neglecting the relationship between rationales and input text. To address this issue, we propose a novel rationalization method that matches the distributions of rationales and input text in both the feature space and output space. Empirically, the proposed distribution matching approach consistently outperforms previous methods by a large margin. Our data and code are available.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Rationalization on HotelReview](<https://www.sota2.com/research/sota/rationalization-on-hotelreview.md>): Rationalization; HotelReview; Precision: 43.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2106.00320>)

[PDF](<https://arxiv.org/pdf/2106.00320.pdf>)
