TY  - RPRT
TI  - Human-Grounded Calibration for Long-Text Image-Text Congruence in Vision-Language Models
AU  - Alessandro Gambetti
AU  - Qiwei Han
PY  - 2026
UR  - https://arxiv.org/abs/2609.15640
ID  - 2609.15640
ER  -