Dobby-Mini-Leashed-Llama-3.1-8B: benchmark results
SentientAGI's persona fine-tune of Llama 3.1 8B Instruct with a blunt, pro-freedom, pro-crypto voice; 'Leashed' is the toned-down variant. Provider: SentientAGI. Released 2025-01-22. Access: Open.
Unified ELO 1502 ± 1, rank #683 of 1392 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 78.47 | Score | 95.1 |
| Open LLM Leaderboard - MATH Level 5 | 18.58 | Score | 68.4 |
| Open LLM Leaderboard - MuSR | 11.94 | Score | 64.4 |
| Open LLM Leaderboard - MMLU-Pro | 29.94 | Score | 59.6 |
| Open LLM Leaderboard - GPQA | 6.94 | Score | 58.3 |
| Open LLM Leaderboard - BBH | 30.77 | Score | 54.4 |
| UGI - Willingness (W/10) | 4.8 | W/10 Score | 36.8 |
| UGI Leaderboard | 27.5 | UGI Score | 24.3 |
| UGI - Natural Intelligence | 13.93 | NatInt Score | 14.2 |
| UGI - Writing | 18.06 | Writing Score | 10 |
Interactive version: theaggregate.ai/model?slug=dobby-mini-leashed-llama-3-1-8b · How It Works · Data refreshed daily, snapshot 2026-09-05.