1 paper
Yue Yao, Zelin Wen, Yan Tong +5
Test-time scaling offers a promising way to improve the reasoning performance of vision-language large models (VLLMs) without additional training. In this paper, we explore a simpl…