llama3-Ko-3-8B-finetuned_ver2: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1420 ± 17, rank #1750 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 27.78 | Normalized accuracy (%) | 95 |
| Open Korean LLM - KorNAT-CKA | 27.46 | Normalized accuracy (%) | 70.2 |
| Open Korean LLM - KorNAT-Harmless | 63.67 | Normalized accuracy (%) | 65.4 |
| Open Korean LLM - Ko-IFEval | 32.39 | Strict accuracy, prompt/instruction mean (%) | 62.3 |
| Open Korean LLM - KorNAT-Helpful | 47.75 | Normalized accuracy (%) | 60 |
| Open Korean LLM - Ko-EQ-Bench | 8.2 | EQ-Bench score | 54.6 |
| Open Korean LLM Leaderboard | 35.16 | Average Score (%) | 50.1 |
| Open Ko-LLM Leaderboard | 35.16 | Average (%) | 50 |
| Open Korean LLM - Ko-Winogrande | 57.77 | Accuracy (%) | 34.3 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 4.78 | Exact match, flexible extract (%) | 23.4 |
Interactive version: theaggregate.ai/model?slug=llama3-ko-3-8b-finetuned-ver2 · How It Works · Data refreshed daily, snapshot 2026-09-19.