Gemma2-9B-IT-Simpo-Infinity-Preference: benchmark results
BAAI's SimPO preference-optimization fine-tune of Gemma 2 9B IT on its Infinity-Preference dataset. Provider: BAAI. Released 2024-08-28. Access: Open.
Aggregate Rating 1520 ± 22, rank #723 of 1654 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-IFEval | 66.22 | Strict accuracy, prompt/instruction mean (%) | 98.4 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 69.07 | Exact match, flexible extract (%) | 97.8 |
| Open Korean LLM - KorNAT-CKA | 35.74 | Normalized accuracy (%) | 95.8 |
| Open Korean LLM - Ko-EQ-Bench | 51.3 | EQ-Bench score | 95.4 |
| Open Korean LLM Leaderboard | 45.48 | Average Score (%) | 92 |
| Open Portuguese LLM - ENEM | 77.19 | Accuracy (%) | 91 |
| Open Portuguese LLM - ASSIN2 RTE | 93.08 | Macro F1 (%) | 86.1 |
| Open Portuguese LLM - BLUEX | 64.81 | Accuracy (%) | 85.7 |
| Open Portuguese LLM - OAB Exams | 53.99 | Accuracy (%) | 84.9 |
| Open LLM Leaderboard - GPQA | 33.98 | Score | 84.4 |
| Open LLM Leaderboard - BBH | 59.79 | Score | 81.5 |
| Open Korean LLM - Ko-Winogrande | 64.09 | Accuracy (%) | 77.1 |
Interactive version: theaggregate.ai/model?slug=gemma2-9b-it-simpo-infinity-preference · How It Works · Data refreshed daily, snapshot 2026-10-24.