llama3.1-korean-v1.1-sft-by-aidx: benchmark results
Provider: Meta. Released 2024-09-29. Access: Open.
Unified ELO 1622 ± 17, rank #545 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 42.06 | Normalized accuracy (%) | 99.6 |
| Open Ko-LLM Leaderboard | 53.94 | Average (%) | 99.2 |
| Open Korean LLM Leaderboard | 53.94 | Average Score (%) | 99.2 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 69.45 | Exact match, flexible extract (%) | 98.1 |
| Open Korean LLM - Ko-IFEval | 62.8 | Strict accuracy, prompt/instruction mean (%) | 97.3 |
| Open Korean LLM - Ko-EQ-Bench | 55.86 | EQ-Bench score | 96.9 |
| Open Korean LLM - Ko-GPQA-Diamond | 27.78 | Normalized accuracy (%) | 95 |
| Open Korean LLM - Ko-Winogrande | 68.75 | Accuracy (%) | 93.9 |
| Open Korean LLM - KorNAT-Harmless | 65.8 | Normalized accuracy (%) | 82.5 |
| Open Korean LLM - KorNAT-Helpful | 48.27 | Normalized accuracy (%) | 66.9 |
Interactive version: theaggregate.ai/model?slug=llama3-1-korean-v1-1-sft-by-aidx · How It Works · Data refreshed daily, snapshot 2026-09-19.