ContraSafe-VL: A Contrastive Safety Preference Dataset for Vision-Language Models
arXiv:2610.23908v1 Announce Type: new Abstract: Vision-Language Models (VLMs) are increasingly deployed as general-purpose multimodal assistants, making multimodal safety alignment an important challenge. Safety depends on the interplay between textual and visual content, with visual context often…
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-10-20 04:00 · arXiv cs.CV
ContraSafe-VL: A Contrastive Safety Preference Dataset for Vision-Language Models