BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors
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_ | 1866913753091538944 |
|---|---|
| author | Hsu, Chia-Yi Tsai, Yu-Lin Zhe, Yu Chen, Yan-Lun Lin, Chih-Hsun Yu, Chia-Mu Zhang, Yang Huang, Chun-Ying Sakuma, Jun |
| author_facet | Hsu, Chia-Yi Tsai, Yu-Lin Zhe, Yu Chen, Yan-Lun Lin, Chih-Hsun Yu, Chia-Mu Zhang, Yang Huang, Chun-Ying Sakuma, Jun |
| contents | Task arithmetic in large-scale pre-trained models enables agile adaptation to diverse downstream tasks without extensive retraining. By leveraging task vectors (TVs), users can perform modular updates through simple arithmetic operations like addition and subtraction. Yet, this flexibility presents new security challenges. In this paper, we investigate how TVs are vulnerable to backdoor attacks, revealing how malicious actors can exploit them to compromise model integrity. By creating composite backdoors that are designed asymmetrically, we introduce BadTV, a backdoor attack specifically crafted to remain effective simultaneously under task learning, forgetting, and analogy operations. Extensive experiments show that BadTV achieves near-perfect attack success rates across diverse scenarios, posing a serious threat to models relying on task arithmetic. We also evaluate current defenses, finding they fail to detect or mitigate BadTV. Our results highlight the urgent need for robust countermeasures to secure TVs in real-world deployments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_02373 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors Hsu, Chia-Yi Tsai, Yu-Lin Zhe, Yu Chen, Yan-Lun Lin, Chih-Hsun Yu, Chia-Mu Zhang, Yang Huang, Chun-Ying Sakuma, Jun Machine Learning Cryptography and Security Task arithmetic in large-scale pre-trained models enables agile adaptation to diverse downstream tasks without extensive retraining. By leveraging task vectors (TVs), users can perform modular updates through simple arithmetic operations like addition and subtraction. Yet, this flexibility presents new security challenges. In this paper, we investigate how TVs are vulnerable to backdoor attacks, revealing how malicious actors can exploit them to compromise model integrity. By creating composite backdoors that are designed asymmetrically, we introduce BadTV, a backdoor attack specifically crafted to remain effective simultaneously under task learning, forgetting, and analogy operations. Extensive experiments show that BadTV achieves near-perfect attack success rates across diverse scenarios, posing a serious threat to models relying on task arithmetic. We also evaluate current defenses, finding they fail to detect or mitigate BadTV. Our results highlight the urgent need for robust countermeasures to secure TVs in real-world deployments. |
| title | BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors |
| topic | Machine Learning Cryptography and Security |
| url | https://arxiv.org/abs/2501.02373 |