1 paper · 1 filter
Runlin Lei, Xiaokui Xiao, Zhewei Wei
Graphs have been used to enhance large language models (LLMs) for structured reasoning, mostly as external knowledge sources are provided to models at test time. In this paper, we…