TY  - RPRT
TI  - Do Vision-Language Models See or Guess? Measuring and Reducing Textual-Prior Reliance with a Phrasing-Controlled Benchmark
AU  - Pratham Singla
AU  - Shivank Garg
AU  - Vihan Singh
AU  - Paras Chopra
PY  - 2026
UR  - https://arxiv.org/abs/2606.10400
ID  - 2606.10400
ER  -