Another experimental LoRA, this has been trained to turn realistic prompts into something like drawings, usually abstract, geometric, and/or primitive. You are more likely to get something like the second image than the first. But with the right prompt, or combined with other LoRAs...?