TY  - RPRT
TI  - Value-Guided Iterative Refinement and the DIQ-H Benchmark for Evaluating VLM Robustness
AU  - Hanwen Wan
AU  - Zexin Lin
AU  - Yixuan Deng
AU  - Xiaoqiang Ji
PY  - 2026
UR  - https://arxiv.org/abs/2512.03992
ID  - 2512.03992
ER  -