I2P

anchor

Findings

• IC-1123All seven published concept erasure methods applied to Stable Diffusion 1.4 can be circumvented by learned word embeddings, demonstrating that targeted concepts are input-filtered rather than truly removed from the model [eval]
• IC-1391SLD concept removal variants and SD with negative prompts are bypassable by Ring-a-Bell adversarial prompts, increasing attack success rate from single digits to 90-100% for nudity [eval]
• IC-445Stable Diffusion v1.5 generates nudity for 796 out of 4703 prompts in the I2P inappropriate prompts dataset [eval]