K2S3-SOLAR-11B-v4.0: benchmark results
K2S3 release published by Changgil: a full-parameter supervised fine-tune of upstage/SOLAR-10.7B-v1.0 using Korean dictionary, KULLM, thesis-abstract and AI Hub data. Its card additionally lists alpaca-gpt4-data and OpenOrca samples. Provider: Changgil. Access: Open.
Aggregate Rating 1422 ± 36, rank #5058 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - WinoGrande | 82.56 | Accuracy (%) (5-shot) | 84.6 |
| Open LLM Leaderboard v1 - MMLU | 65.13 | Accuracy (%) (5-shot) | 82.2 |
| Open LLM Leaderboard v1 - HellaSwag | 84.75 | Normalized accuracy (%) (10-shot) | 68.2 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.65 | Normalized accuracy (%) (25-shot) | 63.7 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 51.63 | MC2 (%) (0-shot) | 52.6 |
| Open LLM Leaderboard v1 - GSM8K | 0 | Accuracy (%) (5-shot) | 4.5 |
Interactive version: theaggregate.ai/model?slug=k2s3-solar-11b-v4-0 · How It Works · Data refreshed daily, snapshot 2026-11-08.