NeuralLLaMa-3-8B-DT-v0.1: benchmark results
Kukedlc's DARE TIES merge of three community Llama 3 8B variants (ChimeraLlama, Stella, and a linear merge). Provider: Kukedlc. Released 2024-05-01. Access: Open.
Aggregate Rating 1462 ± 19, rank #923 of 1656 rated models, from 42 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open CoT - LogiQA | 11.34 | CoT Gain (%) | 97.7 |
| Open LLM Leaderboard v1 - GSM8K | 70.96 | Accuracy (%) (5-shot) | 94.8 |
| Open Medical LLM - PubMedQA | 77.6 | Accuracy (%) | 94 |
| Open CoT - LSAT Logical Reasoning | 19.61 | CoT Gain (%) | 91.6 |
| Open CoT - LogiQA 2 | 14.89 | CoT Gain (%) | 91.6 |
| Open LLM Leaderboard v1 - MMLU | 68.52 | Accuracy (%) (5-shot) | 91.6 |
| Open CoT Leaderboard | 13.33 | Average CoT Gain (%) | 88.5 |
| Open Medical LLM - MedMCQA | 57.73 | Accuracy (%) | 85.7 |
| Open Medical LLM - MedQA (USMLE) | 61.67 | Accuracy (%) | 85.4 |
| Open Medical LLM - MMLU Medical Genetics | 83 | Accuracy (%) | 84.1 |
| Open LLM Leaderboard v1 - ARC Challenge | 69.71 | Normalized accuracy (%) (25-shot) | 84 |
| Open Portuguese LLM - ENEM | 71.24 | Accuracy (%) | 79.8 |
Interactive version: theaggregate.ai/model?slug=neuralllama-3-8b-dt-v0-1 · How It Works · Data refreshed daily, snapshot 2026-10-20.