evaluation-family · Source-linked discovery

Bbeh Mini

A reasoning capability dataset that replaces each task in BIG-Bench-Hard with a novel task that probes a similar reasoning capability but exhibits significantly increased difficulty.

Open interactive record →
OriginMehran Kazemi, Bahare Fatemi, Hritik Bansal 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.

Original sources