Daredevil-8B-abliterated: benchmark results
mlabonne's abliterated version of his Daredevil-8B Llama 3 merge, with the refusal direction removed for uncensored use. Provider: Other. Released 2024-06-01. Access: Open.
Unified ELO 1534 ± 1, rank #518 of 1392 rated models, from 31 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open CoT - LSAT Reading Comprehension | 23.42 | CoT Gain (%) | 96.2 |
| Open Korean LLM Leaderboard | 46.98 | Average Score (%) | 94.3 |
| Open CoT Leaderboard | 14.25 | Average CoT Gain (%) | 93.9 |
| Open CoT - LogiQA | 9.42 | CoT Gain (%) | 93.5 |
| Open Medical LLM - MMLU Medical Genetics | 84 | Accuracy (%) | 93.2 |
| Open CoT - LogiQA 2 | 15.01 | CoT Gain (%) | 92.7 |
| Open Medical LLM - MedMCQA | 58.36 | Accuracy (%) | 89.2 |
| Open Medical LLM - PubMedQA | 77 | Accuracy (%) | 89.2 |
| Open CoT - LSAT Logical Reasoning | 18.63 | CoT Gain (%) | 87.8 |
| Open Medical LLM - MedQA (USMLE) | 61.27 | Accuracy (%) | 84.7 |
| Open Medical LLM | 70.86 | Average Accuracy (%) | 84.1 |
| Open Medical LLM - MMLU College Biology | 79.17 | Accuracy (%) | 80.7 |
Interactive version: theaggregate.ai/model?slug=daredevil-8b-abliterated · How It Works · Data refreshed daily, snapshot 2026-09-05.