Llama-3-Instruct-8B-ORPO: benchmark results
Princeton NLP's Llama-3-Instruct-8B-ORPO, a Llama 3 8B Instruct model aligned via ORPO, released alongside Princeton's SimPO preference-optimization paper. Provider: Princeton NLP. Released 2024-05-17. Access: Open.
Aggregate Rating 1452 ± 19, rank #991 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 26.77 | Normalized accuracy (%) | 88.9 |
| Open Korean LLM - Ko-EQ-Bench | 37.54 | EQ-Bench score | 88.3 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 53.83 | Exact match, flexible extract (%) | 87.5 |
| Open Korean LLM - Ko-IFEval | 42.42 | Strict accuracy, prompt/instruction mean (%) | 86.2 |
| Open LLM Leaderboard - IFEval | 71.28 | Score | 84.6 |
| Open Ko-LLM Leaderboard | 42.03 | Average (%) | 84.5 |
| Open Korean LLM Leaderboard | 42.03 | Average Score (%) | 84.5 |
| Open Portuguese LLM - FaQuAD NLI | 74.92 | Macro F1 (%) | 71.9 |
| Open Portuguese LLM - OAB Exams | 49.29 | Accuracy (%) | 71.9 |
| Open Portuguese LLM - ENEM | 68.72 | Accuracy (%) | 70.8 |
| Open Portuguese LLM - ASSIN2 RTE | 91.17 | Macro F1 (%) | 64.6 |
| Open Portuguese LLM - BLUEX | 54.66 | Accuracy (%) | 60.8 |
Interactive version: theaggregate.ai/model?slug=llama-3-instruct-8b-orpo · How It Works · Data refreshed daily, snapshot 2026-10-23.