Context Quality Matters in Training Fusion-in-Decoder for Extractive Open-Domain Question Answering.

Kosuke Akimoto, Kunihiro Takeoka, Masafumi Oyamada

Browse the full EMNLP paper archive.