Part of my facial expressions series. All of these LoRAs are designed to mesh well with each other and have low impact on your image framing.
This one is interesting to me because it's something that a lot of the base models can do ok with a fresh picture, but it's hard to compose into pictures with other stuff going on. The LoRA seems to be able to do a way better job of that.
The 2.0 version is significantly more face focused and a little stronger than the previous version, but it should also interfere less with the subject. If you want to use it in for the same contexts, you might need to swap the keyword from shocked to surprised or lower the weight.
Starter Prompts:
<subject>, shocked expression
<subject>, BREAK shocked expression