Observations on Muse-Glimmer reasoning traces being noticeably different from qwen / gemma models and questions for you guys

Just downloaded the model, UD-Q5_K_XL quant, asked it to generate a long story to test out reasoning and speed with dflash (super fast btw, ~ 90 to 160 tok/s on a 5090 depending on task) and was surprised at the reasoning traces, they are so unlike anything i've seen recently either in gemma 4, qwen

中文全文翻译正在补齐，可先查看下方英文原文。

Just downloaded the model, UD-Q5_K_XL quant, asked it to generate a long story to test out reasoning and speed with dflash (super fast btw, ~ 90 to 160 tok/s on a 5090 depending on task) and was surprised at the reasoning traces, they are so unlike anything i've seen recently either in gemma 4, qwen