作者位次 #2/3ACL 2026Poster· Natural Language Processing
D ^2 P lan: Dual-Agent Dynamic Global Planning for Complex Retrieval-Augmented Reasoning
Kangcheng Luo · Tinglang Wu · Yansong Feng
摘要
Recent search-augmented LLMs trained with reinforcement learning (RL) can interleave searching and reasoning for multi-hop reasoning tasks. However, they face two critical failure modes as the accumulating context becomes flooded with both crucial evidence and irrelevant information: (1) ineffect...