MMLU-by-task - Formal Logic — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 2 70B Chat | 55.56 |
| 2 | StableBeluga2 | 50 |
| 3 | Llama 2 70B Base | 47.62 |
| 4 | falcon-180B | 47.62 |
| 5 | LLaMA-65B | 43.65 |
| 6 | Llama 2 70B Chat (HF) | 41.27 |
| 7 | Mistral-7B-v0.1 | 41.27 |
| 8 | Llama 2 70B Chat GPTQ | 40.48 |
| 9 | WizardCoder-Python-34B-V1.0 | 38.89 |
| 10 | internlm-20B Chat | 38.89 |
| 11 | vicuna-7B-v1.5 | 38.1 |
| 12 | vicuna-13B-v1.3 | 37.3 |
| 13 | vicuna-33B-v1.3 | 37.3 |
| 14 | vicuna-13B-v1.5 | 37.3 |
| 15 | trurl-2-7B | 36.51 |
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-formal-logic · How the rankings work · Data refreshed daily, snapshot 2026-07-22.