DeepSeek-R1-Distill-Qwen-32B-abliterated: benchmark results
huihui-ai's abliterated derivative of deepseek-ai/DeepSeek-R1-Distill-Qwen-32B, described by its publisher as modifying refusal behavior. It remains distinct from the upstream release. Provider: huihui-ai. Access: Open.
Aggregate Rating 1501 ± 35, rank #2334 of 10916 rated models, from 50 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Japanese LLM - Xlsum JA Rouge1 | 33.22 | Score (%) | 91.3 |
| Open Japanese LLM - Xlsum JA Bert Score JA F1 | 70.76 | Score (%) | 90.7 |
| Open Japanese LLM - SUM | 11.74 | Score (%) | 90.6 |
| Open Japanese LLM - Xlsum JA Rouge2 | 11.74 | Score (%) | 90.6 |
| Open Japanese LLM - Xlsum JA RougeLsum | 26 | Score (%) | 85.6 |
| Open Japanese LLM - Wiki NER SET F1 | 9.73 | Score (%) | 76.1 |
| Open Japanese LLM - Jsts Pearson | 89.69 | Score (%) | 73.8 |
| Open Japanese LLM - Jsick Exact Match | 82.14 | Score (%) | 73.2 |
| Open Japanese LLM - Jsem Exact Match | 78.47 | Score (%) | 68.5 |
| Open Japanese LLM - Jnli Exact Match | 83.94 | Score (%) | 63.7 |
| Open Japanese LLM - Jsts Spearman | 86.41 | Score (%) | 60.6 |
| Open Japanese LLM - NLI | 75.93 | Score (%) | 54.9 |
Interactive version: theaggregate.ai/model?slug=deepseek-r1-distill-qwen-32b-abliterated · How It Works · Data refreshed daily, snapshot 2026-11-08.