List your company
ResearchFormerly WizWand

Reading Comprehension on DROP 1.0 (test)

92.38EMHuman Performance
0.65224.46648.2872.094Aug 15, 2019Aug 17, 2019Aug 20, 2019Aug 23, 2019Aug 25, 2019Aug 28, 2019Aug 31, 2019
Updated 8mo ago

Evaluation results

MethodLinks
Human Performance2019.0892.3895.98
Giving BERT a CalculatorOperations=All, Pre-training=CoQA, Ensemble=6 models2019.0878.1481.78
Giving BERT a CalculatorOperations=All, Pre-training=CoQA2019.0876.9680.53
MTMSNModel Scale=Large2019.0875.8579.88
NAQANetModel=NAQANet2019.0844.2447.77
NAQANet2019.0844.0747.01
BERTModel Scale=Base2019.0829.4532.7
QANet+ELMo2019.0827.0829.67
BiDAF2019.0824.7527.49
Semantic Role Labeling2019.0810.8713.35
Heuristic Baseline2019.084.188.59