Year: 2025
Type: preprint
Abstract: Parallel thinking expands exploration breadth, complementing the deep exploration of information-seeking (IS) agents to further enhance problem-solving capability. However, conventional parallel thinking faces two key challenges in this setting: inefficiency from repeatedly rolling out from scratch, and difficulty in integrating long-horizon reasoning trajectories during answer generation, as limited context capacity prevents full consideration of the reasoning process. To address these issues, we propose ParallelMuse, a two-stage paradigm designed for deep IS agents. The first stage, Functionality-Specified Partial Rollout, partitions generated sequences into functional regions and performs uncertainty-guided path reuse and branching to enhance exploration efficiency. The second stage, Compressed Reasoning Aggregation, exploits reasoning redundancy to losslessly compress information relevant to answer derivation and synthesize a coherent final answer. Experiments across multiple open-source agents and benchmarks demonstrate up to 62% performance improvement with a 10--30% reduction in exploratory token consumption.
Source: arXiv (Cornell University)
Authors: Li, Baixuan, Zhang, Dingchu, Wu, Jialong, Yin, Wenbiao, Tao, Zhengwei, Zhao, Yida, Zhang, Liwen, Shen, Haiyang, Fang, Runnan, Xie, Pengjun, Zhou, Jingren, Jiang, Yong
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7102401851
Cites:
Cited by:
Related to: