Select High-quality Synthetic QA Pairs to Augment Training Data in MRC under the Reward Guidance of Generative Language Models.

Jing Jin, Houfeng Wang

Browse the full COLING paper archive.