ppo2: benchmark results
Provider: Other. Access: Open.
Unified ELO 1397 ± 18, rank #1930 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-CKA | 33.14 | Normalized accuracy (%) | 90.8 |
| Open Korean LLM - Ko-GPQA-Diamond | 25.25 | Normalized accuracy (%) | 75.9 |
| Open Korean LLM - Ko-IFEval | 27.6 | Strict accuracy, prompt/instruction mean (%) | 44.2 |
| Open Korean LLM - Ko-Winogrande | 58.88 | Accuracy (%) | 42.2 |
| Open Korean LLM Leaderboard | 33.41 | Average Score (%) | 41.6 |
| Open Ko-LLM Leaderboard | 33.41 | Average (%) | 41.3 |
| Open Korean LLM - KorNAT-Harmless | 61.76 | Normalized accuracy (%) | 33.5 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - KorNAT-Helpful | 43 | Normalized accuracy (%) | 20.7 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 1.36 | Exact match, flexible extract (%) | 13.1 |
Interactive version: theaggregate.ai/model?slug=ppo2 · How It Works · Data refreshed daily, snapshot 2026-09-19.