TY  - RPRT
TI  - GeoPID: Decomposing and Steering Visual Information in Vision-Language Models
AU  - Seulgi Kim
AU  - Zhixiong Zhang
AU  - Xinwei Zhang
AU  - Jie Ling
AU  - Ronn Shaw
PY  - 2026
UR  - https://arxiv.org/abs/2610.08401
ID  - 2610.08401
ER  -