MMLU-by-task - Philosophy — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 1257 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Llama 2 70B Base | 78.78 |
| 2 | falcon-180B | 78.78 |
| 3 | Llama 2 70B Chat | 75.88 |
| 4 | StableBeluga2 | 75.56 |
| 5 | LLaMA-65B | 72.67 |
| 6 | Llama 2 70B Chat GPTQ | 71.7 |
| 7 | Llama 2 70B Chat (HF) | 70.42 |
| 8 | Mistral-7B-v0.1 | 69.77 |
| 9 | vicuna-33B-v1.3 | 68.17 |
| 10 | LLaMA-30B | 66.56 |
| 11 | MythoMax-L2-13B | 64.31 |
| 12 | falcon-40B Instruct | 63.99 |
| 13 | vicuna-13B-v1.5 | 63.99 |
| 14 | internlm-20B Chat | 63.34 |
| 15 | OpenHermes-13B | 60.77 |
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-philosophy · How the rankings work · Data refreshed daily, snapshot 2026-07-22.