Qwen2.5-14B-Instruct-abliterated-v2: benchmark results
huihui-ai's uncensored community variant of Qwen2.5-14B-Instruct with refusal behavior removed via abliteration. Provider: Alibaba. Released 2024-10-09.
Aggregate Rating 1560 ± 52, rank #1451 of 10870 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 83.28 | Score | 99.3 |
| Open LLM Leaderboard - MATH Level 5 | 53.02 | Score | 98.2 |
| Open LLM Leaderboard - MMLU-Pro | 49.62 | Score | 87.5 |
| UGI - Willingness (W/10) | 8.2 | W/10 Score | 87.2 |
| Open LLM Leaderboard - BBH | 63.24 | Score | 87.1 |
| Open LLM Leaderboard - GPQA | 33.39 | Score | 82 |
| Open LLM Leaderboard - MuSR | 42.2 | Score | 62.6 |
| UGI Leaderboard | 38.84 | UGI Score | 62 |
| UGI - Natural Intelligence | 19.21 | NatInt Score | 36.8 |
| UGI - Writing | 29.12 | Writing Score | 33 |
Interactive version: theaggregate.ai/model?slug=qwen2-5-14b-instruct-abliterated-v2 · How It Works · Data refreshed daily, snapshot 2026-11-12.