Can I ask your opinion?

I am always trying new things when generating images, experimenting to see what works best in different situations. For instance, I find that some styles work better for some characters than others, and some handle certain poses much better than others. And I like trying new models for characters, poses, concepts, etc. But every model, LoRa, embedding, etc that I add (or take away) impacts the overall look and feel of an image.

Here are three images of Nelson, numbered for your convenience. They all use khanykov01's character model, and the style models I most commonly use, but they have different combinations of negative embeddings (or none at all). Don't worry if you don't know what negative embeddings are -- I just want your opinion on which image you like best.

Plus, this is an excuse for me to try using the poll function :) So, please let me know if you have a preference for Image #1, Image #2, or Image #3. Maybe you don't see a difference between them? Or maybe you don't care and are happy with any one of them? Thanks very much!