ARC-Challenge, ARC-Easy, BoolQ, HellaSwag (HS), OpenBookQA (OBQA), PIQA, Winogrande (WG), MMLU
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
ARC-Challenge, ARC-Easy, BoolQ, HellaSwag (HS), OpenBookQA (OBQA), PIQA, Winogrande (WG), MMLU (test)
53.16ARC-Challenge Accuracy
9