Tag
multimodal
2 articles

Research/Apr 15
Rubric-Based DPO for Visual Preference Tuning
rDPO uses instance-specific rubrics to make visual preference optimization more fine-grained, improving filtering and benchmark results.

Research/Apr 3
Steerable ViT Features for Text-Guided Vision
A new vision representation lets text steer ViT features toward specific objects without giving up generic visual utility.