K-EXAONE-236B-A23B: benchmark results
Provider: LG AI. Access: Open.
Unified ELO 1557 ± 1, rank #413 of 1392 rated models, from 37 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Nejumi 4 - ALT - Toxicity | 84.17 | Score (%) | 81.1 |
| Nejumi 4 - ALT - Bias | 89.47 | Score (%) | 70.1 |
| LLM Stats Score | 26.13 | LLM Stats Score (conservative rating) | 59.8 |
| OpenEvals - MMLU-Pro | 83.8 | Accuracy (%) | 59.4 |
| Nejumi 4 - ALT - Truthfulness | 74.12 | Score (%) | 52.5 |
| Nejumi 4 - GLP - Specialized Knowledge | 59.94 | Score (%) | 47.5 |
| Nejumi 4 - GLP - Logical Reasoning | 86 | Score (%) | 42.6 |
| OpenEvals - GPQA Diamond | 79.1 | Accuracy (%) | 42.3 |
| Nejumi 4 - ALT Average | 83.66 | Score (%) | 41 |
| Nejumi 4 - GLP - Mathematical Reasoning | 93.17 | Score (%) | 40.6 |
| LLM Stats (MMMLU) | 85.7 | Score (%) | 40 |
| Nejumi 4 - ALT - Controllability | 85.18 | Score (%) | 39.3 |
Interactive version: theaggregate.ai/model?slug=k-exaone-236b-a23b · How It Works · Data refreshed daily, snapshot 2026-09-05.