NeuralMonarch-7B: benchmark results
mlabonne's NeuralMonarch 7B, a DPO fine-tune of his Monarch-7B merge using the truthy-dpo and Intel Orca DPO preference datasets. Provider: mlabonne. Released 2024-03-01. Access: Open.
Aggregate Rating 1455 ± 15, rank #944 of 1635 rated models, from 40 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM - KorNAT-Harmless | 79.7 | Normalized accuracy (%) | 98.9 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 77.79 | MC2 (%) (0-shot) | 97.3 |
| Open LLM Leaderboard v1 - HellaSwag | 89.09 | Normalized accuracy (%) (10-shot) | 96.9 |
| Open LLM Leaderboard v1 - ARC Challenge | 73.21 | Normalized accuracy (%) (25-shot) | 96.3 |
| Open LLM Leaderboard v1 - WinoGrande | 84.61 | Accuracy (%) (5-shot) | 94 |
| Open Korean LLM - KorNAT-Helpful | 52.35 | Normalized accuracy (%) | 92.6 |
| Open Chinese LLM - TruthfulQA MC | 62.76 | Accuracy (%) | 92.3 |
| YALL Nous Leaderboard | 57.53 | Average | 92.1 |
| Open Korean LLM - KorNAT-CKA | 32.82 | Normalized accuracy (%) | 89.5 |
| Open Korean LLM Leaderboard | 42.37 | Average Score (%) | 86.3 |
| Open LLM Leaderboard v1 - GSM8K | 67.78 | Accuracy (%) (5-shot) | 84.4 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 43.06 | Exact match, flexible extract (%) | 78.5 |
Interactive version: theaggregate.ai/model?slug=neuralmonarch-7b · How It Works · Data refreshed daily, snapshot 2026-11-05.