A novel interface for adversarial trivia question-writing

arXiv:2404.00011v1 Announce Type: cross
Abstract: A critical component when developing question-answering AIs is an adversarial dataset that challenges models to adapt to the complex syntax and reasoning underlying our natural language. Present techniques for procedurally generating adversarial texts are not robust enough for training on complex tasks such as answering multi-sentence trivia questions. We instead turn to human-generated data by introducing an interface for collecting adversarial human-written trivia questions. Our interface is aimed towards question writers and players of Quiz …

abstract adapt adversarial arxiv challenges cs.cl cs.hc dataset language natural natural language novel question questions reasoning robust syntax tasks training type writing