Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2410.00296 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910626770583552 |
|---|---|
| author | Du, Xuefeng Ghosh, Reshmi Sim, Robert Salem, Ahmed Carvalho, Vitor Lawton, Emily Li, Yixuan Stokes, Jack W. |
| author_facet | Du, Xuefeng Ghosh, Reshmi Sim, Robert Salem, Ahmed Carvalho, Vitor Lawton, Emily Li, Yixuan Stokes, Jack W. |
| contents | Vision-language models (VLMs) are essential for contextual understanding of both visual and textual information. However, their vulnerability to adversarially manipulated inputs presents significant risks, leading to compromised outputs and raising concerns about the reliability in VLM-integrated applications. Detecting these malicious prompts is thus crucial for maintaining trust in VLM generations. A major challenge in developing a safeguarding prompt classifier is the lack of a large amount of labeled benign and malicious data. To address the issue, we introduce VLMGuard, a novel learning framework that leverages the unlabeled user prompts in the wild for malicious prompt detection. These unlabeled prompts, which naturally arise when VLMs are deployed in the open world, consist of both benign and malicious information. To harness the unlabeled data, we present an automated maliciousness estimation score for distinguishing between benign and malicious samples within this unlabeled mixture, thereby enabling the training of a binary prompt classifier on top. Notably, our framework does not require extra human annotations, offering strong flexibility and practicality for real-world applications. Extensive experiment shows VLMGuard achieves superior detection results, significantly outperforming state-of-the-art methods. Disclaimer: This paper may contain offensive examples; reader discretion is advised. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_00296 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | VLMGuard: Defending VLMs against Malicious Prompts via Unlabeled Data Du, Xuefeng Ghosh, Reshmi Sim, Robert Salem, Ahmed Carvalho, Vitor Lawton, Emily Li, Yixuan Stokes, Jack W. Machine Learning Cryptography and Security Vision-language models (VLMs) are essential for contextual understanding of both visual and textual information. However, their vulnerability to adversarially manipulated inputs presents significant risks, leading to compromised outputs and raising concerns about the reliability in VLM-integrated applications. Detecting these malicious prompts is thus crucial for maintaining trust in VLM generations. A major challenge in developing a safeguarding prompt classifier is the lack of a large amount of labeled benign and malicious data. To address the issue, we introduce VLMGuard, a novel learning framework that leverages the unlabeled user prompts in the wild for malicious prompt detection. These unlabeled prompts, which naturally arise when VLMs are deployed in the open world, consist of both benign and malicious information. To harness the unlabeled data, we present an automated maliciousness estimation score for distinguishing between benign and malicious samples within this unlabeled mixture, thereby enabling the training of a binary prompt classifier on top. Notably, our framework does not require extra human annotations, offering strong flexibility and practicality for real-world applications. Extensive experiment shows VLMGuard achieves superior detection results, significantly outperforming state-of-the-art methods. Disclaimer: This paper may contain offensive examples; reader discretion is advised. |
| title | VLMGuard: Defending VLMs against Malicious Prompts via Unlabeled Data |
| topic | Machine Learning Cryptography and Security |
| url | https://arxiv.org/abs/2410.00296 |