How to turn trivia questions into state-of-the-art AI training data [Research, EMNLP 2024]

Presentation of our EMNLP 2024 paper: You Make me Feel like a Natural Question: Training QA Systems on Transformed Trivia Questions Many of the questions for training AIs how to answer questions come from the queries users type into search engines (like Google's Natural Questions). Is there a cheaper---perhaps even better---way? We propose a "naturalization" technique to turn high-quality, rigorously edited trivia questions into examples that resembles Natural Questions. Training on our naturalized questions and testing on natural questions comes close to the results with using Natural Questions, and we can improve results on MMLU (a standard modern evaluation set) by using our data. Code and Data: https://github.com/Pinafore/qb2nq Full paper: