HELM Classic - WikiFact: leaderboard
The "WikiFact" subtask of HELM Classic. A scenario from Stanford HELM Classic, the holistic evaluation suite scoring language models across accuracy, robustness and calibration on a broad set of tasks.
Metric: Exact Match (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 2 70B | 44.92 | #1201 |
| 2 | LLaMA-65B | 42.08 | #1303 |
| 3 | text-davinci-002 | 39.25 | #1225 |
| 4 | falcon-40B | 38.03 | #1442 |
| 5 | text-davinci-003 | 37.34 | #1146 |
| 6 | mpt-30B | 36.93 | #1476 |
| 7 | Llama 2 13B | 36.47 | #1436 |
| 8 | falcon-40B Instruct | 35.92 | #1458 |
| 9 | Mistral-7B-v0.1 | 34.87 | #1326 |
| 10 | Palmyra X (43B) | 33.82 | #1317 |
| 11 | Llama 2 7B | 33.49 | #1520 |
| 12 | mpt-30B Instruct | 32.76 | #1470 |
| 13 | falcon-7B | 30.87 | #1598 |
| 14 | davinci | 30.61 | #1553 |
| 15 | GPT-3.5 Turbo (0613) | 28.86 | #1118 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-classic-wikifact · How It Works · Data refreshed daily, snapshot 2026-10-20.