Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.23573 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915698555486208 |
|---|---|
| author | Yu, Shaohan Li, Lijun Si, Chenyang Sheng, Lu Shao, Jing |
| author_facet | Yu, Shaohan Li, Lijun Si, Chenyang Sheng, Lu Shao, Jing |
| contents | The rapid evolution of generative models has led to a continuous emergence of multimodal safety risks, exposing the limitations of existing defense methods. To address these challenges, we propose ProGuard, a vision-language proactive guard that identifies and describes out-of-distribution (OOD) safety risks without the need for model adjustments required by traditional reactive approaches. We first construct a modality-balanced dataset of 87K samples, each annotated with both binary safety labels and risk categories under a hierarchical multimodal safety taxonomy, effectively mitigating modality bias and ensuring consistent moderation across text, image, and text-image inputs. Based on this dataset, we train our vision-language base model purely through reinforcement learning (RL) to achieve efficient and concise reasoning. To approximate proactive safety scenarios in a controlled setting, we further introduce an OOD safety category inference task and augment the RL objective with a synonym-bank-based similarity reward that encourages the model to generate concise descriptions for unseen unsafe categories. Experimental results show that ProGuard achieves performance comparable to closed-source large models on binary safety classification, substantially outperforms existing open-source guard models on unsafe content categorization. Most notably, ProGuard delivers a strong proactive moderation ability, improving OOD risk detection by 52.6% and OOD risk description by 64.8%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_23573 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ProGuard: Towards Proactive Multimodal Safeguard Yu, Shaohan Li, Lijun Si, Chenyang Sheng, Lu Shao, Jing Computer Vision and Pattern Recognition The rapid evolution of generative models has led to a continuous emergence of multimodal safety risks, exposing the limitations of existing defense methods. To address these challenges, we propose ProGuard, a vision-language proactive guard that identifies and describes out-of-distribution (OOD) safety risks without the need for model adjustments required by traditional reactive approaches. We first construct a modality-balanced dataset of 87K samples, each annotated with both binary safety labels and risk categories under a hierarchical multimodal safety taxonomy, effectively mitigating modality bias and ensuring consistent moderation across text, image, and text-image inputs. Based on this dataset, we train our vision-language base model purely through reinforcement learning (RL) to achieve efficient and concise reasoning. To approximate proactive safety scenarios in a controlled setting, we further introduce an OOD safety category inference task and augment the RL objective with a synonym-bank-based similarity reward that encourages the model to generate concise descriptions for unseen unsafe categories. Experimental results show that ProGuard achieves performance comparable to closed-source large models on binary safety classification, substantially outperforms existing open-source guard models on unsafe content categorization. Most notably, ProGuard delivers a strong proactive moderation ability, improving OOD risk detection by 52.6% and OOD risk description by 64.8%. |
| title | ProGuard: Towards Proactive Multimodal Safeguard |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.23573 |