HumanLikeness - Word-2: leaderboard
HumanLikeness word-level task comparing model responses with human word-meaning priming patterns.
Metric: Humanlike Score (%). Source: huggingface.co. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | zephyr-7B-beta | 25.64 | #1330 |
| 2 | Llama 2 13B Chat Base | 25.05 | #1422 |
| 3 | Mistral 7B Instruct (v0.3) | 25 | #1274 |
| 4 | starchat2-15B-v0.1 | 24.52 | #832 |
| 5 | CodeLlama-34B-Instruct-hf | 23.32 | #1381 |
| 6 | zephyr-7B-alpha | 22.88 | #1312 |
| 7 | Mixtral 8x7B Instruct (v0.1) | 22.7 | #1073 |
| 8 | Llama 2 7B Chat | 22.13 | #1499 |
| 9 | Phi-3 Mini 4K Instruct | 19.26 | #1297 |
| 10 | Llama 3 8B Instruct | 14.47 | #1129 |
| 11 | Mistral 7B Instruct (v0.2) | 13.73 | #1333 |
| 12 | Yi 1.5 34B Chat | 12.63 | #800 |
| 13 | Llama 3.1 8B Instruct | 11.81 | #1035 |
| 14 | Mistral Nemo Instruct (2407) | 9.86 | #936 |
| 15 | c4ai-command-r-plus | 8.24 | #724 |
Interactive version: theaggregate.ai/benchmark?slug=humanlikeness-word-2 · How It Works · Data refreshed daily, snapshot 2026-10-20.