free-solar-evo-v0.1: benchmark results
Freewheelin AI's free-solar-evo v0.1, built on hwkwon's S-SOLAR-10.7B-v1.5 base using a method inspired by a Sakana AI project. Provider: Freewheelin. Released 2024-04-18. Access: Open.
Aggregate Rating 1421 ± 19, rank #1140 of 1656 rated models, from 29 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-Helpful | 77.53 | Normalized accuracy (%) | 99.7 |
| Open Korean LLM - KorNAT-Harmless | 83.83 | Normalized accuracy (%) | 99.5 |
| Open Chinese LLM - TruthfulQA MC | 68.79 | Accuracy (%) | 98.8 |
| Open Korean LLM - Ko-Winogrande | 73.01 | Accuracy (%) | 98.5 |
| Open LLM Leaderboard - MuSR | 49.46 | Score | 97.6 |
| Open Korean LLM - KorNAT-CKA | 34.69 | Normalized accuracy (%) | 94.3 |
| Open Chinese LLM - HellaSwag | 67.99 | Accuracy (%) | 85.9 |
| Open Korean LLM - Ko-GPQA-Diamond | 26.26 | Normalized accuracy (%) | 85.5 |
| Open Korean LLM Leaderboard | 40.39 | Average Score (%) | 76.9 |
| Open Ko-LLM Leaderboard | 40.39 | Average (%) | 76.8 |
| Open Chinese LLM - WinoGrande | 65.9 | Accuracy (%) | 75.3 |
| Open Chinese LLM - ARC Challenge | 54.95 | Accuracy (%) | 66.9 |
Interactive version: theaggregate.ai/model?slug=free-solar-evo-v0-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.