unlearning-or-concealment

Public

We expose a significant vulnerability in diffusion model unlearning methods, where an attacker can reverse the supposed erasure of concepts during the inference process. Our approach leverages a novel Partial Diffusion Attack that operates across all layers of the model.

diffusion diffusion-models evaluation evaluation-metrics

Creat：2024-08-23T11:37:34

Update：2025-03-07T12:17:03

https://respailab.github.io/unlearning-or-concealment

Stars

Stars Increase

Related projects

Stable Diffusion Webui

Stable Diffusion web UI

157286

1年前

+26today

Transformers

Hot

bert

? Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

151173

2年前

+68today

Comfyui

Hot

ComfyUI docker images for use in GPU cloud and local environments. Includes AI-Dock base for authentication and improved user experience.

91095

6个月前

+114today

Django

apps

The Web framework for perfectionists with deadlines.

85423

6个月前

+14today

LLMs From Scratch

Hot

Implement a ChatGPT-like LLM in PyTorch from scratch, step by step

75539

1年前

+96today

Gpt_academic

academic

为GPT/GLM等LLM大语言模型提供实用化交互接口，特别优化论文阅读/润色/写作体验，模块化设计，支持自定义快捷按钮&函数插件，支持Python和C++等项目剖析&自译解功能，PDF/LaTex论文翻译&总结功能，支持并行问询多种LLM模型，支持chatglm3等本地模型。接入通义千问, deepseekcoder, 讯飞星火, 文心一言, llama2, rwkv, claude2, moss等。

69343

6个月前

+13today