evaluation-task · Source-linked discovery
BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions
Reading comprehension dataset that queries for complex, non-factoid information, and require difficult entailment-like inference to solve.
OriginChristopher Clark, Kenton Lee, Ming-Wei Chang et al.TopicsGeneral capabilityStatusimported
Can support
Not independently assessed by FronteraEval yet.
Cannot support by itself
No inference beyond the upstream source should be made until the protocol is reviewed.