Qwen/Qwen2-72B
| benchmark | category | score | context | conf | evidence |
|---|---|---|---|---|---|
| HellaSwag | knowledge | 87.8 | HellaSwag | medium | | HellaSwag | 87.8 | **88.7** | 88.0 | 86.0 | 87.5 | 87.6 | |
| CMMLU | knowledge | 84.0 | CMMLU | medium | | CMMLU | 84.0 | 53.4 | 67.2 | 83.5 | 88.3 | **90.1** | |
| C-Eval | knowledge | 81.7 | C-Eval | medium | | C-Eval | 81.7 | 54.6 | 65.2 | 84.1 | 89.1 | **91.0** | |
| GSM8K | math | 79.2 | GSM8K | medium | | GSM8K | 79.2 | 83.7 | 83.0 | 79.5 | 85.4 | **89.5** | |
| BBH | reasoning | 78.9 | BBH | medium | | BBH | 78.9 | 78.9 | 81.0 | 65.5 | 74.8 | **82.4** | |
| MMLU | knowledge | 78.5 | MMLU | medium | | MMLU | 78.5 | 77.8 | 79.5 | 77.5 | 80.4 | **84.2** | |
| MBPP | coding | 73.9 | MBPP | medium | | MBPP | 73.9 | 71.7 | 70.4 | 66.9 | 70.9 | **76.9** | |
| EvalPlus | coding | 55.0 | EvalPlus | medium | | EvalPlus | 55.0 | 54.1 | 54.8 | 52.9 | 57.7 | **65.4** | |
| MMLU-Pro | knowledge | 49.5 | MMLU-Pro | medium | | MMLU-Pro | - | 49.5 | 52.8 | 45.8 | 49.4 | **55.6** | |
| HumanEval | coding | 45.7 | HumanEval | medium | | HumanEval | 45.7 | 46.3 | 48.2 | 46.3 | 54.3 | **64.6** | |
| MATH | math | 43.6 | MATH | medium | | MATH | 43.6 | 41.7 | 42.5 | 34.1 | 49.6 | **51.1** | |
| TruthfulQA | safety | 42.2 | TruthfulQA | medium | | TruthfulQA | 42.2 | 51.0 | 45.6 | **59.6** | 49.6 | 54.8 | |
| GPQA | reasoning | 34.3 | GPQA | medium | | GPQA | - | 34.3 | 36.3 | 36.3 | 35.9 | **37.9** | |
none scraped.