LLM benchmarks
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
10 LLM Benchmarks suite (AIME25, AIME26, CEval, GPQA-D, HLE, IMO-AB, MMLU, MMLU-Pro, MMLU-Rd, MultiCh)
70AIME 25 Score
20
LLM Benchmarks (ARC, HellaSwag, MMLU, TruthfulQA, Winogrande, GSM8K)
55.55ARC Score
9
42 LLM benchmarks Aggregate (overall)
67.14Average Score
5