1 paper · 1 filter
Jinwei Yao, Kaiqi Chen, Kexun Zhang +4
Large language models (LLMs) are increasingly employed for complex tasks that process multiple generation calls in a tree structure with shared prefixes of tokens, including few-sh…