Eukaryote-8x7B-bf16: benchmark results
Kquant03's mixture-of-experts release whose published Mixtral configuration specifies 8 local experts, 2 selected per token and 32 decoder layers. Its recipe names SanjiWatsuki/Kunoichi-DPO-v2-7B as the base; nominal expert sizes are not measured total or active parameter counts. Provider: Kquant03. Access: Open.
Aggregate Rating 1452 ± 37, rank #3684 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - GSM8K | 71.95 | Accuracy (%) (5-shot) | 97 |
| Open LLM Leaderboard v1 - HellaSwag | 87.29 | Normalized accuracy (%) (10-shot) | 85.3 |
| Open LLM Leaderboard v1 - WinoGrande | 82.4 | Accuracy (%) (5-shot) | 83.9 |
| Open LLM Leaderboard v1 - ARC Challenge | 69.45 | Normalized accuracy (%) (25-shot) | 83.2 |
| Open LLM Leaderboard v1 - MMLU | 65.15 | Accuracy (%) (5-shot) | 82.3 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 63.17 | MC2 (%) (0-shot) | 80.2 |
Interactive version: theaggregate.ai/model?slug=eukaryote-8x7b-bf16 · How It Works · Data refreshed daily, snapshot 2026-11-08.