papers AI Learner
The Github is limit! Click to go to the new site.

Trick Me If You Can: Human-in-the-loop Generation of Adversarial Examples for Question Answering

2019-05-15
Eric Wallace, Pedro Rodriguez, Shi Feng, Ikuya Yamada, Jordan Boyd-Graber

Abstract

Adversarial evaluation stress tests a model’s understanding of natural language. While past approaches expose superficial patterns, the resulting adversarial examples are limited in complexity and diversity. We propose human-in-the-loop adversarial generation, where human authors are guided to break models. We aid the authors with interpretations of model predictions through an interactive user interface. We apply this generation framework to a question answering task called Quizbowl, where trivia enthusiasts craft adversarial questions. The resulting questions are validated via live human–computer matches: although the questions appear ordinary to humans, they systematically stump neural and information retrieval models. The adversarial questions cover diverse phenomena from multi-hop reasoning to entity type distractors, exposing open challenges in robust question answering.

Abstract (translated by Google)
URL

http://arxiv.org/abs/1809.02701

PDF

http://arxiv.org/pdf/1809.02701


Comments

Content