Saved in:
Bibliographic Details
Main Authors: Zhang, Shuo, Yuan, Chaofa, Guo, Ryan, Yu, Xiaomin, Xu, Rui, Chen, Zhangquan, Li, Zinuo, Yang, Zhi, Guan, Shuhao, Tang, Zhenheng, Hu, Sen, Zhang, Liwen, Chen, Ronghao, Wang, Huacan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.09465
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914305368129536
author Zhang, Shuo
Yuan, Chaofa
Guo, Ryan
Yu, Xiaomin
Xu, Rui
Chen, Zhangquan
Li, Zinuo
Yang, Zhi
Guan, Shuhao
Tang, Zhenheng
Hu, Sen
Zhang, Liwen
Chen, Ronghao
Wang, Huacan
author_facet Zhang, Shuo
Yuan, Chaofa
Guo, Ryan
Yu, Xiaomin
Xu, Rui
Chen, Zhangquan
Li, Zinuo
Yang, Zhi
Guan, Shuhao
Tang, Zhenheng
Hu, Sen
Zhang, Liwen
Chen, Ronghao
Wang, Huacan
contents While LLM-based agents have shown promise for deep research, most existing approaches rely on fixed workflows that struggle to adapt to real-world, open-ended queries. Recent work therefore explores self-evolution by allowing agents to rewrite their own code or prompts to improve problem-solving ability, but unconstrained optimization often triggers instability, hallucinations, and instruction drift. We propose EvoFSM, a structured self-evolving framework that achieves both adaptability and control by evolving an explicit Finite State Machine (FSM) instead of relying on free-form rewriting. EvoFSM decouples the optimization space into macroscopic Flow (state-transition logic) and microscopic Skill (state-specific behaviors), enabling targeted improvements under clear behavioral boundaries. Guided by a critic mechanism, EvoFSM refines the FSM through a small set of constrained operations, and further incorporates a self-evolving memory that distills successful trajectories as reusable priors and failure patterns as constraints for future queries. Extensive evaluations on five multi-hop QA benchmarks demonstrate the effectiveness of EvoFSM. In particular, EvoFSM reaches 58.0% accuracy on the DeepSearch benchmark. Additional results on interactive decision-making tasks further validate its generalization.
format Preprint
id arxiv_https___arxiv_org_abs_2601_09465
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
Zhang, Shuo
Yuan, Chaofa
Guo, Ryan
Yu, Xiaomin
Xu, Rui
Chen, Zhangquan
Li, Zinuo
Yang, Zhi
Guan, Shuhao
Tang, Zhenheng
Hu, Sen
Zhang, Liwen
Chen, Ronghao
Wang, Huacan
Artificial Intelligence
While LLM-based agents have shown promise for deep research, most existing approaches rely on fixed workflows that struggle to adapt to real-world, open-ended queries. Recent work therefore explores self-evolution by allowing agents to rewrite their own code or prompts to improve problem-solving ability, but unconstrained optimization often triggers instability, hallucinations, and instruction drift. We propose EvoFSM, a structured self-evolving framework that achieves both adaptability and control by evolving an explicit Finite State Machine (FSM) instead of relying on free-form rewriting. EvoFSM decouples the optimization space into macroscopic Flow (state-transition logic) and microscopic Skill (state-specific behaviors), enabling targeted improvements under clear behavioral boundaries. Guided by a critic mechanism, EvoFSM refines the FSM through a small set of constrained operations, and further incorporates a self-evolving memory that distills successful trajectories as reusable priors and failure patterns as constraints for future queries. Extensive evaluations on five multi-hop QA benchmarks demonstrate the effectiveness of EvoFSM. In particular, EvoFSM reaches 58.0% accuracy on the DeepSearch benchmark. Additional results on interactive decision-making tasks further validate its generalization.
title EvoFSM: Controllable Self-Evolution for Deep Research with Finite State Machines
topic Artificial Intelligence
url https://arxiv.org/abs/2601.09465