llama2-ko-13B-instruct: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1311 ± 17, rank #2419 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-Winogrande | 59.98 | Accuracy (%) | 50.5 |
| Open Korean LLM - KorNAT-Helpful | 46.05 | Normalized accuracy (%) | 41.8 |
| Open Korean LLM - KorNAT-Harmless | 61.39 | Normalized accuracy (%) | 27.7 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 1.14 | Exact match, flexible extract (%) | 12.3 |
| Open Korean LLM - KorNAT-CKA | 16.23 | Normalized accuracy (%) | 11.9 |
| Open Korean LLM Leaderboard | 28.98 | Average Score (%) | 9 |
| Open Ko-LLM Leaderboard | 28.98 | Average (%) | 8.9 |
| Open Korean LLM - Ko-IFEval | 6.87 | Strict accuracy, prompt/instruction mean (%) | 6.6 |
| Open Korean LLM - Ko-GPQA-Diamond | 19.19 | Normalized accuracy (%) | 6.4 |
Interactive version: theaggregate.ai/model?slug=llama2-ko-13b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-19.