General Capability Suite
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
General Capability Suite MMLU, AlpacaEval, GSM8K, MATH, HumanEval
71.86MMLU
56
General Capability Suite MMLU, GSM8K, HumanEval, IFEval
77.78Common Average Score
39
General Capability Suite ARC-C, HellaSwag, MMLU, GSM8K
54.27ARC-C Accuracy
27
General Capability Suite HS WG IFEval MMLU
17.7HS Delta
22
General Capability Suite Aggregate
62.51General Capability Avg. Accuracy
18
General Capability Suite (MMLU, TruthfulQA, HellaSwag, ARC-Easy) (test)
0.082MMLU Score
16
General Capability Suite
71Average Score
12
General Capability Suite (MMLU, GSM8K, GPQA)
73.6MMLU Accuracy
5