一作ACL 2026Poster· Natural Language Processing
Revisit What You See: Revealing Visual Semantics in Vision Tokens to Guide LVLM Decoding
Beomsik Cho · Jaehyung Kim
摘要
Large Vision–Language Models (LVLMs) achieve strong performance across multimodal tasks by integrating visual perception with language understanding. However, how vision information contributes to the model’s decoding process remains under-explored, as reflected in frequent hallucinations. Throug...