Hermes-3-Llama-3.1-70B-lorablated — benchmark results
mlabonne's uncensored 'lorablated' variant of Nous Research's Hermes 3 Llama 3.1 70B, with refusals removed via a LoRA abliteration merge. Provider: Nous Research. Released 2024-08-16. Access: Open.
Unified ELO 1566 ± 44, rank #576 of 1776 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 24.73 | Score | 99.2 |
| Open LLM Leaderboard - BBH | 52.75 | Score | 95.6 |
| Open LLM Leaderboard - GPQA | 15.44 | Score | 91.8 |
| Open LLM Leaderboard - MMLU-Pro | 40.88 | Score | 85.8 |
| Open LLM Leaderboard - MATH Level 5 | 22.43 | Score | 75.5 |
| UGI - Natural Intelligence | 29.83 | NatInt Score | 74.1 |
| UGI Leaderboard | 41.46 | UGI Score | 71 |
| UGI - Writing | 37.81 | Writing Score | 65.7 |
| UGI - Willingness (W/10) | 6 | W/10 Score | 53.7 |
| Open LLM Leaderboard - IFEval | 34.24 | Score | 32.5 |
Interactive version: theaggregate.ai/model?slug=hermes-3-llama-3-1-70b-lorablated · How the rankings work · Data refreshed daily, snapshot 2026-07-22.