Finding Generalizable Evidence by Learning to Convince Q&A Models

TL;DR

This paper introduces a system that learns to select the most convincing evidence for answers in passage-based QA, improving model robustness and human answerability with minimal passage content.

Contribution

It presents a novel evidence selection method that generalizes across models and enhances QA performance with less passage information.

Findings

Agent-selected evidence increases answer plausibility across models and humans.

Humans can answer questions correctly with only about 20% of the passage.

QA models generalize better to longer passages and more difficult questions.

Abstract

We propose a system that finds the strongest supporting evidence for a given answer to a question, using passage-based question-answering (QA) as a testbed. We train evidence agents to select the passage sentences that most convince a pretrained QA model of a given answer, if the QA model received those sentences instead of the full passage. Rather than finding evidence that convinces one model alone, we find that agents select evidence that generalizes; agent-chosen evidence increases the plausibility of the supported answer, as judged by other QA models and humans. Given its general nature, this approach improves QA in a robust manner: using agent-selected evidence (i) humans can correctly answer questions with only ~20% of the full passage and (ii) QA models can generalize to longer passages and harder questions.