Question Answering on CNN (val)
79.2AccuracyGA + MAGE
Evaluation results
| Method | Links | |
|---|---|---|
| GA + MAGEBase Architecture=GA, Coreference augmentation=MAGE, Coreference state size=162017.03 | 79.2 | |
| GA + MAGEBase Architecture=GA, Coreference augmentation=MAGE, Coreference state size=322017.03 | 79.2 | |
| GA + one-hotBase Architecture=GA, Coreference augmentation=one-hot2017.03 | 78 | |
| GABase Architecture=GA2017.03 | 77.9 | |
| BiDAFBase Architecture=BiDAF2017.03 | 76.3 | |
| BiGRU + MAGEBase Architecture=BiGRU, Coreference augmentation=MAGE, Coreference state size=482017.03 | 72.2 | |
| BiGRUBase Architecture=BiGRU2017.03 | 69.4 | |
| BiGRU + one-hotBase Architecture=BiGRU, Coreference augmentation=one-hot2017.03 | 69.2 |