作者位次 #2/10ACL 2026Findings· Findings of ACL
A uto S earch: Adaptive Search Depth for Efficient Agentic RAG via Reinforcement Learning
Jingbo Sun · Wenyue Chong · Songjun Tu · Qichao Zhang · Yaocheng Zhang · Jiajun Chai · Xiaohan Wang · Shuai Chen · Guojun Yin · Dongbin Zhao
摘要
Agentic retrieval-augmented generation (RAG) systems enable large language models (LLMs) to solve complex tasks through multi-step interaction with external retrieval tools. However, such multi-step interaction often involves redundant search steps, incurring substantial computational cost and la...