K2S3-SOLAR-11B-v1.0: benchmark results
K2S3 release published by Changgil: a full-parameter supervised fine-tune of upstage/SOLAR-10.7B-Instruct-v1.0 using Korean dictionary, KULLM, thesis-abstract and AI Hub data. Provider: Changgil. Access: Open.
Aggregate Rating 1198 ± 35, rank #9630 of 10916 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 23.23 | Normalized accuracy (%) | 48.1 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 45.99 | MC2 (%) (0-shot) | 33.1 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-IFEval | 22.09 | Strict accuracy, prompt/instruction mean (%) | 22.5 |
| Open Korean LLM - KorNAT-Helpful | 43.23 | Normalized accuracy (%) | 22.4 |
| Open LLM Leaderboard v1 - GSM8K | 1.36 | Accuracy (%) (5-shot) | 16.6 |
| Open LLM Leaderboard v1 - MMLU | 30.05 | Accuracy (%) (5-shot) | 16.6 |
| Open Korean LLM Leaderboard | 29.86 | Average Score (%) | 14.4 |
| Open Korean LLM - Ko-Winogrande | 53.35 | Accuracy (%) | 12.3 |
| Open LLM Leaderboard v1 - ARC Challenge | 33.7 | Normalized accuracy (%) (25-shot) | 11.7 |
| Open Korean LLM - KorNAT-CKA | 15.9 | Normalized accuracy (%) | 11.1 |
| Open LLM Leaderboard v1 - WinoGrande | 57.54 | Accuracy (%) (5-shot) | 10.7 |
Interactive version: theaggregate.ai/model?slug=k2s3-solar-11b-v1-0 · How It Works · Data refreshed daily, snapshot 2026-11-08.