Standard Downstream Tasks
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Standard Downstream Tasks (arc-c, arc-e, boolq, hellaswag, piqa, siqa, winogrande)
70.65ARC-C Accuracy
55
Standard Downstream Tasks PIQA, HellaSwag, Winogrande, ARC-Challenge, ARC-Easy
77.47PIQA Zero-Shot Accuracy
9
Standard Downstream Tasks (ARC, COPA, BoolQ, PIQA, StoryCloze, RTE, MMLU)
49.57ARC (Challenge)
8