Because it is based on LoRA but is trained with Conv2D

Main tag : gusset,
Secondary tag : pantyhose, thighband pantyhose, panties under pantyhose, no panties,
Color tag : color pantyhose, color panties,
Pose tag : from behind, from below, ass, sitting, standing,
LoRA str : 1
Denoising str : <0.4

Conv2D requires need extensions
https://github.com/KohakuBlueleaf/a1111-sd-webui-lycoris
or
https://github.com/kohya-ss/sd-webui-additional-networks