ZIUM: Zero-Shot Intent-Aware Adversarial Attack on Unlearned Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918107244658688 |
|---|---|
| author | Yook, Hyun Jun Jhun, Ga San Cho, Jae Hyun Jeon, Min Kim, Donghyun Kim, Tae Hyung Lee, Youn Kyu |
| author_facet | Yook, Hyun Jun Jhun, Ga San Cho, Jae Hyun Jeon, Min Kim, Donghyun Kim, Tae Hyung Lee, Youn Kyu |
| contents | Machine unlearning (MU) removes specific data points or concepts from deep learning models to enhance privacy and prevent sensitive content generation. Adversarial prompts can exploit unlearned models to generate content containing removed concepts, posing a significant security risk. However, existing adversarial attack methods still face challenges in generating content that aligns with an attacker's intent while incurring high computational costs to identify successful prompts. To address these challenges, we propose ZIUM, a Zero-shot Intent-aware adversarial attack on Unlearned Models, which enables the flexible customization of target attack images to reflect an attacker's intent. Additionally, ZIUM supports zero-shot adversarial attacks without requiring further optimization for previously attacked unlearned concepts. The evaluation across various MU scenarios demonstrated ZIUM's effectiveness in successfully customizing content based on user-intent prompts while achieving a superior attack success rate compared to existing methods. Moreover, its zero-shot adversarial attack significantly reduces the attack time for previously attacked unlearned concepts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_21985 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ZIUM: Zero-Shot Intent-Aware Adversarial Attack on Unlearned Models Yook, Hyun Jun Jhun, Ga San Cho, Jae Hyun Jeon, Min Kim, Donghyun Kim, Tae Hyung Lee, Youn Kyu Computer Vision and Pattern Recognition Cryptography and Security Machine unlearning (MU) removes specific data points or concepts from deep learning models to enhance privacy and prevent sensitive content generation. Adversarial prompts can exploit unlearned models to generate content containing removed concepts, posing a significant security risk. However, existing adversarial attack methods still face challenges in generating content that aligns with an attacker's intent while incurring high computational costs to identify successful prompts. To address these challenges, we propose ZIUM, a Zero-shot Intent-aware adversarial attack on Unlearned Models, which enables the flexible customization of target attack images to reflect an attacker's intent. Additionally, ZIUM supports zero-shot adversarial attacks without requiring further optimization for previously attacked unlearned concepts. The evaluation across various MU scenarios demonstrated ZIUM's effectiveness in successfully customizing content based on user-intent prompts while achieving a superior attack success rate compared to existing methods. Moreover, its zero-shot adversarial attack significantly reduces the attack time for previously attacked unlearned concepts. |
| title | ZIUM: Zero-Shot Intent-Aware Adversarial Attack on Unlearned Models |
| topic | Computer Vision and Pattern Recognition Cryptography and Security |
| url | https://arxiv.org/abs/2507.21985 |