Gemma-2-9B-It-SPPO-Iter2: benchmark results
UCLA-AGI's second-iteration SPPO self-play preference-optimization research checkpoint of Gemma 2 9B It. Provider: UCLA. Released 2024-06-29. Access: Open.
Aggregate Rating 1547 ± 20, rank #625 of 1661 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-EQ-Bench | 61.4 | EQ-Bench score | 99.6 |
| Open Korean LLM - Ko-IFEval | 66.79 | Strict accuracy, prompt/instruction mean (%) | 98.6 |
| Open Ko-LLM Leaderboard | 50.5 | Average (%) | 97.4 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 68.76 | Exact match, flexible extract (%) | 97.4 |
| Open Korean LLM Leaderboard | 50.5 | Average Score (%) | 97.4 |
| Open Korean LLM - KorNAT-CKA | 34.8 | Normalized accuracy (%) | 94.8 |
| Open Portuguese LLM - ASSIN2 RTE | 94.07 | Macro F1 (%) | 94.4 |
| Open Korean LLM - KorNAT-Harmless | 67.09 | Normalized accuracy (%) | 87.9 |
| Open Korean LLM - KorNAT-Helpful | 50.85 | Normalized accuracy (%) | 87.5 |
| Open Portuguese LLM - ENEM | 73.69 | Accuracy (%) | 85.7 |
| Open Portuguese LLM - OAB Exams | 53.12 | Accuracy (%) | 83 |
| Open Portuguese LLM - BLUEX | 63 | Accuracy (%) | 82.9 |
Interactive version: theaggregate.ai/model?slug=gemma-2-9b-it-sppo-iter2 · How It Works · Data refreshed daily, snapshot 2026-10-23.