Announcement_9

Paper titled “S2H-DPO: Hardness-Aware Preference Optimization for Vision–Language Models” has been accepted to ACL 2026!. 🎊