Few-Shot Out-of-Domain Transfer Learning of Natural Language Explanations in a Label-Abundant Setup

TL;DR

This paper explores methods for transferring natural language explanations from a resource-rich parent task to a low-resource child task, improving explainability in NLP with minimal additional data.

Contribution

It introduces four transfer learning methods for few-shot out-of-domain NLE transfer, demonstrating their effectiveness across multiple NLP tasks.

Findings

Parent task transfer improves NLE quality

Best transfer methods identified for different setups

Effective explanation generation with minimal NLE data

Abstract

Training a model to provide natural language explanations (NLEs) for its predictions usually requires the acquisition of task-specific NLEs, which is time- and resource-consuming. A potential solution is the few-shot out-of-domain transfer of NLEs from a parent task with many NLEs to a child task. In this work, we examine the setup in which the child task has few NLEs but abundant labels. We establish four few-shot transfer learning methods that cover the possible fine-tuning combinations of the labels and NLEs for the parent and child tasks. We transfer explainability from a large natural language inference dataset (e-SNLI) separately to two child tasks: (1) hard cases of pronoun resolution, where we introduce the small-e-WinoGrande dataset of NLEs on top of the WinoGrande dataset, and (2)~commonsense validation (ComVE). Our results demonstrate that the parent task helps with NLE generation and we establish the best methods for this setup.