Llama-3-Instruct-8B-SPPO-Iter3: benchmark results
Provider: Meta. Released 2024-06-25. Access: Open.
Unified ELO 1495 ± 1, rank #715 of 1392 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 68.34 | Score | 81 |
| Open Chinese LLM - GSM8K | 52.08 | Accuracy (%) | 71.2 |
| Open Chinese LLM - TruthfulQA MC | 56.29 | Accuracy (%) | 68.8 |
| Open LLM Leaderboard - MMLU-Pro | 29.53 | Score | 57.3 |
| Open LLM Leaderboard - BBH | 29.74 | Score | 50.7 |
| Open Chinese LLM Leaderboard | 56.62 | Average Score (%) | 50 |
| Open Korean LLM Leaderboard | 35.63 | Average Score (%) | 49.7 |
| Open Chinese LLM - C-Eval Semantic | 68.99 | Accuracy (%) | 47.9 |
| Open Chinese LLM - CMMLU | 53.18 | Accuracy (%) | 47.8 |
| Open LLM Leaderboard - MATH Level 5 | 9.59 | Score | 45.9 |
| Open Chinese LLM - ARC Challenge | 49.57 | Accuracy (%) | 39.6 |
| Ru Arena Hard | 47.45 | Win Rate (%) | 38.5 |
Interactive version: theaggregate.ai/model?slug=llama-3-instruct-8b-sppo-iter3 · How It Works · Data refreshed daily, snapshot 2026-09-05.