llama2-merge-v0.1: benchmark results
Provider: Meta. Access: Open.
Unified ELO 1416 ± 17, rank #1786 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-Winogrande | 64.96 | Accuracy (%) | 81.7 |
| Open Korean LLM - KorNAT-Helpful | 47.95 | Normalized accuracy (%) | 62.9 |
| Open Korean LLM - Ko-IFEval | 30.42 | Strict accuracy, prompt/instruction mean (%) | 56 |
| Open Korean LLM - KorNAT-Harmless | 63.06 | Normalized accuracy (%) | 54.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 18.65 | Exact match, flexible extract (%) | 49.9 |
| Open Korean LLM Leaderboard | 34.86 | Average Score (%) | 49.1 |
| Open Ko-LLM Leaderboard | 34.86 | Average (%) | 49 |
| Open Korean LLM - KorNAT-CKA | 21.17 | Normalized accuracy (%) | 33.1 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-GPQA-Diamond | 18.69 | Normalized accuracy (%) | 4.1 |
Interactive version: theaggregate.ai/model?slug=llama2-merge-v0-1 · How It Works · Data refreshed daily, snapshot 2026-09-19.