Mistral-7B-Instruct-RDPO: benchmark results
Princeton NLP's 7B Mistral Instruct checkpoint tuned with RDPO preference optimization, released as part of their SimPO preference-optimization research line. Provider: Princeton NLP. Released 2024-05-17. Access: Open.
Aggregate Rating 1361 ± 19, rank #1365 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-Harmless | 66.22 | Normalized accuracy (%) | 84.9 |
| Open Korean LLM - Ko-GPQA-Diamond | 25.25 | Normalized accuracy (%) | 75.9 |
| Open Korean LLM - KorNAT-Helpful | 48.15 | Normalized accuracy (%) | 65.1 |
| Open Korean LLM - Ko-IFEval | 32.72 | Strict accuracy, prompt/instruction mean (%) | 63.6 |
| Open LLM Leaderboard - IFEval | 48.87 | Score | 57 |
| Open Portuguese LLM - ASSIN2 RTE | 90.36 | Macro F1 (%) | 56.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 22.82 | Exact match, flexible extract (%) | 55.5 |
| Open Portuguese LLM - FaQuAD NLI | 67.64 | Macro F1 (%) | 54.7 |
| Open Korean LLM - Ko-EQ-Bench | 7.53 | EQ-Bench score | 53.6 |
| Open Portuguese LLM - BLUEX | 50.63 | Accuracy (%) | 47.8 |
| Open Portuguese LLM - ENEM | 59.62 | Accuracy (%) | 45.8 |
| Open Korean LLM Leaderboard | 33.26 | Average Score (%) | 40.3 |
Interactive version: theaggregate.ai/model?slug=mistral-7b-instruct-rdpo · How It Works · Data refreshed daily, snapshot 2026-10-23.