UTF-8 and characters with diacritics, like Czech

[pesaak-githup]

System Info / 系統信息

Hi, I test GLM5 in Kilo code. I like it. But I have a big issue with characters with diacritics. Czech ones in my case.

This is what GLM5 says after we investigate it.
Most likely cause: The issue is in the model's tokenization/output layer for Unicode characters. The GLM-5 model appears to handle Czech characters inconsistently - sometimes corrupting them during output transmission.

I attach the handoff file for another AI model that I must use because GLM5 can not handle it in a good way.
I also attach some screenshots from the thread.

I also noticed that AI somehow cycled when trying to work with notes files and UTF. It tooks like couple of minutes to come back to me, and context grew up very quickly.

I hope you can solve it in next release.

handoff-utf8-fix-prompt.md

Who can help? / 谁可以帮助到您？

No response

Information / 问题信息

• The official example scripts / 官方的示例脚本
• My own modified scripts / 我自己修改的脚本和任务

Reproduction / 复现过程

Just try to use UTF characters in XAML UX files.

Expected behavior / 期待表现

I would appreciate if you can solve it in next release, otherwise I am practically not able to use the model.