List your company

ResearchFormerly WizWand

MultiSpanQA

Benchmarks

TaskDatasetSOTA resultTrendResults

Question Answering

MultiSpanQA (test)

71.4Exact F1

11

Multi-answer Machine Reading Comprehension

MultiSpanQA (val)

0.719EM F1 Score

8

Instance-level Evaluation

MultiSpanQA

72.9AUC-ROC

7

Closed book Question Answering

MultiSpanQA (test)

48F1 Score

5