LoRA to help steer image output for table grinding/masturbation. V1 does not seem to work for a "from behind" view. Works well with character LoRAs from what I saw while testing, but watch your weights as I noticed if the character LoRA weight is not reduced slightly or you use 1.0 or higher on this LoRA you can get bad hands.Quick test showed that the LoRA drops off quickly around 0.5. Recommended weight range in this case looks be 0.8 ~ 1.2.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/dea1aaf6-fbde-4519-afc9-d5f9f6710e37/width=525/dea1aaf6-fbde-4519-afc9-d5f9f6710e37.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/95d18205-81ed-46ad-98b0-580c7d4a2066/width=525/95d18205-81ed-46ad-98b0-580c7d4a2066.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/7b3a429f-5757-40bf-b479-f7b7b1be7852/width=525/7b3a429f-5757-40bf-b479-f7b7b1be7852.jpeg" />

Table Masturbation Concept Augment

Initial release. From behind does not replicate intent for LoRA when I tested it. Might work with some prompts, but the ones I used did not output what I was going for. 1.0 weight worked well. Recommend lowering to 0.8 - 0.9 if you are using character LoRAs. Note: I only tested combined with my own character LoRA.

Files with SHA256: cc773547f1fd4e321edfdd156fedb031488b046df1f129f9b49780793795184d