Mixolar-4x7B: benchmark results
shadowml's mergekit MoE of four SOLAR-10.7B fine-tunes (Sakura-SOLAR, CarbonVillain, meow, SOLRCA-Math) despite the 4x7B name. Provider: shadowml. Released 2023-12-30. Access: Open.
Aggregate Rating 1456 ± 22, rank #964 of 1656 rated models, from 25 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 71.81 | MC2 (%) (0-shot) | 92.3 |
| Open LLM Leaderboard v1 - HellaSwag | 88.44 | Normalized accuracy (%) (10-shot) | 91.8 |
| Open LLM Leaderboard v1 - WinoGrande | 83.58 | Accuracy (%) (5-shot) | 89.7 |
| Open LLM Leaderboard v1 - ARC Challenge | 71.08 | Normalized accuracy (%) (25-shot) | 87.9 |
| Open LLM Leaderboard v1 - MMLU | 66.29 | Accuracy (%) (5-shot) | 86.6 |
| Open Chinese LLM - TruthfulQA MC | 61.09 | Accuracy (%) | 86.4 |
| Open Portuguese LLM - ASSIN2 RTE | 92.93 | Macro F1 (%) | 85 |
| Open LLM Leaderboard v1 - GSM8K | 63.91 | Accuracy (%) (5-shot) | 77.6 |
| Open Portuguese LLM - FaQuAD NLI | 75.81 | Macro F1 (%) | 74.6 |
| Open Portuguese LLM - BLUEX | 58.55 | Accuracy (%) | 73.6 |
| Open Portuguese LLM - ENEM | 68.93 | Accuracy (%) | 71.5 |
| Open LLM Leaderboard - MuSR | 42.58 | Score | 66.2 |
Interactive version: theaggregate.ai/model?slug=mixolar-4x7b · How It Works · Data refreshed daily, snapshot 2026-10-20.