Your favorite fastest abliterated/safety removed 3.6 and 3.8 27b?

查看原文

Qwen模型发布订阅权益

safety

热度趋势

新上榜

最近 24 小时与此前 24 小时对比 · 7 天曲线

百分比基于当前可用热度信号，而非评论数或独立用户人数。

推荐理由

Qwen 相关模型动态已经出现，适合跟踪能力变化、生态影响和后续可用性。

AI 摘要

A Reddit user is seeking to compare abliterated (safety-removed) Qwen 3.6 27b and Qwen 3.8 27b models for low-thinking tasks. They plan to use instruction-following benchmarks with "thinking off" to determine which version performs better. The user notes that abliterated models are preferred for their specific needs, and currently uses "Huihui-Qwen3.8-27B-abliterated-NVFP4-GGUF" based on community comments.