HELM Classic - bAbI: leaderboard
The "bAbI" 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. 70 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 2 70B | 70.53 | #1201 |
| 2 | Palmyra X (43B) | 70.24 | #1317 |
| 3 | LLaMA-65B | 67.66 | #1303 |
| 4 | text-davinci-003 | 65.34 | #1146 |
| 5 | LLaMA-30B | 63.36 | #1423 |
| 6 | text-davinci-002 | 61.81 | #1225 |
| 7 | falcon-40B Instruct | 60.6 | #1458 |
| 8 | Mistral-7B-v0.1 | 59.27 | #1326 |
| 9 | Llama 2 13B | 58.39 | #1436 |
| 10 | falcon-40B | 57.62 | #1442 |
| 11 | LLaMA-13B | 56.07 | #1516 |
| 12 | Llama 2 7B | 54.97 | #1520 |
| 13 | mpt-30B | 54.97 | #1476 |
| 14 | RedPajama-INCITE-7B-Instruct | 54.53 | #1566 |
| 15 | vicuna-13B-v1.3 | 53.31 | #1447 |
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-babi · How It Works · Data refreshed daily, snapshot 2026-10-20.