Jun 2026
ADAPT: Attention Dynamics Alignment with Preference Tuning for Faithful MLLMs
This work provides an attention-based perspective on mitigating hallucinations by exploring the model's internal text-to-image cross-attention behaviors, and proposes ADAPT, an attention-based framework that intervenes directly on text-to-image cross-attention dynamics.
Zhiyuan Yao, Zheren Fu, Zhi Zheng et al.
· arXiv.org · 0 citations