Llama-3-Instruct-8B-SPPO-Iter1: benchmark results
Provider: Meta. Released 2024-06-25. Access: Open.
Aggregate Rating 1462 ± 38, rank #3205 of 10916 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-EQ-Bench | 39.65 | EQ-Bench score | 90.8 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 56.03 | Exact match, flexible extract (%) | 90.1 |
| Open Korean LLM - Ko-IFEval | 43 | Strict accuracy, prompt/instruction mean (%) | 87.4 |
| Open LLM Leaderboard - IFEval | 72.99 | Score | 87.2 |
| Open Korean LLM - KorNAT-Helpful | 49.68 | Normalized accuracy (%) | 80.7 |
| Open Korean LLM Leaderboard | 40.91 | Average Score (%) | 79.7 |
| Open LLM Leaderboard - MMLU-Pro | 37.11 | Score | 60.6 |
| Open Korean LLM - Ko-GPQA-Diamond | 23.74 | Normalized accuracy (%) | 55.6 |
| Open Korean LLM - Ko-Winogrande | 60.46 | Accuracy (%) | 54.5 |
| Open Korean LLM - KorNAT-Harmless | 62.93 | Normalized accuracy (%) | 52.2 |
| Open LLM Leaderboard - MATH Level 5 | 11.48 | Score | 52 |
| Open LLM Leaderboard - BBH | 50.58 | Score | 50.9 |
Interactive version: theaggregate.ai/model?slug=llama-3-instruct-8b-sppo-iter1 · How It Works · Data refreshed daily, snapshot 2026-11-08.