1 paper · 1 filter
Xinglong Yang, Zhilin Peng, Zhanzhan Liu +2
Self-consistency methods are the core technique for improving the reasoning reliability of multimodal large language models (MLLMs). By generating multiple reasoning results throug…