lowreactivity/style

Experimental style LoRA

This LoRA was created as a personal experiment for my own workflow.

What it does:
It suppresses contrast and saturation, producing a grayer, dustier, more matte render.
I made it to counter overly bright and glossy outputs that persisted even with restrained prompts.

Known issue:
This LoRA currently struggles with backgrounds. In some cases, it significantly flattens or removes background detail, turning it into near–solid color fields.
This is not an intentional stylistic choice, but a limitation of the training and weighting.

As a result, the LoRA performs best on portraits and close-ups, where the background is less critical.

The recommended strength is approximately 0.7–0.9, though lower values may help preserve background structure.

Trigger word: lowreact