Mistral-Small-Instruct-2409-abliterated: benchmark results
byroneverson's abliterated variant of Mistral Small Instruct 2409 (22B) with refusal behavior removed. Provider: Mistral. Released 2024-09-23. Access: Open.
Unified ELO 1517 ± 1, rank #619 of 1392 rated models, from 15 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Korean LLM Leaderboard | 47.12 | Average Score (%) | 94.7 |
| Open Chinese LLM - GSM8K | 59.51 | Accuracy (%) | 83.4 |
| Open LLM Leaderboard - IFEval | 69.71 | Score | 82.7 |
| Open LLM Leaderboard - GPQA | 11.07 | Score | 81.7 |
| Open LLM Leaderboard - MATH Level 5 | 24.77 | Score | 77.9 |
| Open Chinese LLM Leaderboard | 60.5 | Average Score (%) | 71.8 |
| Open LLM Leaderboard - MMLU-Pro | 32.48 | Score | 71.3 |
| Open Chinese LLM - WinoGrande | 64.8 | Accuracy (%) | 69 |
| Open Chinese LLM - C-Eval Semantic | 74.44 | Accuracy (%) | 63.8 |
| Open Chinese LLM - HellaSwag | 61.38 | Accuracy (%) | 62.8 |
| Open Chinese LLM - CMMLU | 56.71 | Accuracy (%) | 62.5 |
| Open Chinese LLM - ARC Challenge | 53.5 | Accuracy (%) | 56.7 |
Interactive version: theaggregate.ai/model?slug=mistral-small-instruct-2409-abliterated · How It Works · Data refreshed daily, snapshot 2026-09-05.