MMLU-by-task - HellaSwag: leaderboard
The "HellaSwag" subtask of MMLU-by-task. A single MMLU subject, reporting per-task multiple-choice accuracy from the MMLU-by-task leaderboard across a wide range of academic and professional domains.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 1257 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Xwin-LM-70B-V0.1 | 68.72 | #982 |
| 2 | Llama 2 70B Chat | 68.09 | #1237 |
| 3 | Llama 2 70B Base | 67.61 | #1025 |
| 4 | StableBeluga2 | 67.45 | #1078 |
| 5 | Llama 2 70B Chat (HF) | 66.94 | #1264 |
| 6 | WizardLM-70B-V1.0 | 65.35 | #1236 |
| 7 | mpt-30B Instruct | 65.25 | #1470 |
| 8 | falcon-40B Instruct | 64.62 | #1458 |
| 9 | MythoMax-L2-13B | 64.23 | #1409 |
| 10 | MPT-30B-Chat | 63.53 | #1451 |
| 11 | vicuna-33B-v1.3 | 63.15 | #1347 |
| 12 | Mistral-7B-v0.1 | 62.95 | #1326 |
| 13 | WizardLM-13B-V1.2 | 62.95 | #1425 |
| 14 | Llama 2 13B Chat Base | 62.94 | #1422 |
| 15 | OpenHermes-13B | 62.47 | #1393 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mmlu-by-task-hellaswag · How It Works · Data refreshed daily, snapshot 2026-10-20.