# Explain Yourself! Leveraging Language Models for Commonsense Reasoning

## TL;DR

This paper introduces CoS-E, a new dataset of human explanations for commonsense reasoning, and proposes CAGE, a framework that uses language models to generate explanations, significantly improving performance on the CommonsenseQA task.

## Contribution

The paper presents a novel dataset of human explanations and a framework that leverages language models to generate explanations, enhancing commonsense reasoning in neural networks.

## Key findings

- CAGE improves state-of-the-art by 10% on CommonsenseQA.
- Language models can effectively generate explanations for reasoning tasks.
- Explanations transfer to out-of-domain tasks, aiding generalization.

## Abstract

Deep learning models perform poorly on tasks that require commonsense reasoning, which often necessitates some form of world-knowledge or reasoning over information not immediately present in the input. We collect human explanations for commonsense reasoning in the form of natural language sequences and highlighted annotations in a new dataset called Common Sense Explanations (CoS-E). We use CoS-E to train language models to automatically generate explanations that can be used during training and inference in a novel Commonsense Auto-Generated Explanation (CAGE) framework. CAGE improves the state-of-the-art by 10% on the challenging CommonsenseQA task. We further study commonsense reasoning in DNNs using both human and auto-generated explanations including transfer to out-of-domain tasks. Empirical results indicate that we can effectively leverage language models for commonsense reasoning.