POLAR-14B-v0.1: benchmark results
Provider: Other. Access: Open.
Unified ELO 1471 ± 18, rank #1367 of 2656 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - Ko-GPQA-Diamond | 31.82 | Normalized accuracy (%) | 100 |
| Open Korean LLM - KorNAT-Harmless | 83.03 | Normalized accuracy (%) | 99.4 |
| Open Korean LLM - KorNAT-Helpful | 74 | Normalized accuracy (%) | 99.4 |
| Open Korean LLM - KorNAT-CKA | 34.8 | Normalized accuracy (%) | 94.8 |
| Open Korean LLM - Ko-Winogrande | 68.19 | Accuracy (%) | 92.6 |
| Open Korean LLM Leaderboard | 39.49 | Average Score (%) | 72.4 |
| Open Ko-LLM Leaderboard | 39.49 | Average (%) | 72.1 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 9.63 | Exact match, flexible extract (%) | 33 |
| Open Korean LLM - Ko-EQ-Bench | 0 | EQ-Bench score | 24.5 |
| Open Korean LLM - Ko-IFEval | 18.1 | Strict accuracy, prompt/instruction mean (%) | 13.4 |
Interactive version: theaggregate.ai/model?slug=polar-14b-v0-1 · How It Works · Data refreshed daily, snapshot 2026-09-19.