llama3-koen-8B-sft-v1: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1464 ± 17, rank #1416 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 32.39 | Normalized accuracy (%) | 88.8 |
| Open Korean LLM - Ko-Winogrande | 62.9 | Accuracy (%) | 71.1 |
| Open Korean LLM - Ko-EQ-Bench | 25.22 | EQ-Bench score | 71 |
| Open Korean LLM - Ko-GPQA-Diamond | 24.75 | Normalized accuracy (%) | 69.5 |
| Open Ko-LLM Leaderboard | 38.53 | Average (%) | 67.7 |
| Open Korean LLM Leaderboard | 38.54 | Average Score (%) | 67.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 27.52 | Exact match, flexible extract (%) | 61.7 |
| Open Korean LLM - Ko-IFEval | 28.97 | Strict accuracy, prompt/instruction mean (%) | 50.1 |
| Open Korean LLM - KorNAT-Helpful | 45.85 | Normalized accuracy (%) | 39.8 |
| Open Korean LLM - KorNAT-Harmless | 61.52 | Normalized accuracy (%) | 30.4 |
Interactive version: theaggregate.ai/model?slug=llama3-koen-8b-sft-v1 · How It Works · Data refreshed daily, snapshot 2026-09-19.