ACL Findings2026
SGA-MCTS: Decoupling Planning from Execution via Training-Free Atomic Experience Retrieval
Xin Xie, Dongyun Xue, Wuguannan Yao, Mingxiao Feng, Wengang Zhou, Xiang Qi, Houqiang Li, Peng Zhang
Abstract
LLM-powered systems require complex multistep decision-making abilities to solve realworld tasks, yet current planning approaches face a trade-off between the high latency of inference-time search and the limited generalization of supervised fine-tuning. To address this limitation, we introduce SGA-MCTS, a framework that casts LLM planning as nonparametric retrieval. Offline, we leverage Monte Carlo Tree Search (MCTS) to explore the solution space and distill high-fidelity trajectories into State-Goal-Action (SGA) atoms. These atoms are de-lexicalized primitives that abstract concrete entities into symbolic slots, preserving reusable causal logic while discarding domain-specific noise. Online, a retrievalaugmented agent employs a hybrid symbolicsemantic mechanism to fetch relevant SGAs and re-ground them into the current context as soft reasoning hints. Empirical results on complex benchmarks demonstrate that this paradigm enables frozen, open-weights models to match the performance of SOTA systems (e.g., GPT-5) without task-specific fine-tuning. By effectively amortizing the heavy computational cost of search, SGA-MCTS achieves System 2 reasoning depth at System 1 inference speeds, rendering autonomous planning both scalable and real-time feasible.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Reflexion: language agents with verbal reinforcement learning cites Toolformer: Language Models Can Teach Themselves to Use Tools
- ExpeL: LLM Agents Are Experiential Learners cites Reflexion: language agents with verbal reinforcement learning
- ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs cites Reflexion: language agents with verbal reinforcement learning
- ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs cites Toolformer: Language Models Can Teach Themselves to Use Tools
Builds on23
- Toolformer: Language Models Can Teach Themselves to Use ToolsTimo Schick, Jane Dwivedi-Yu, Roberto Dessì, Roberta Raileanu et al.NeurIPS 2023 · 5,989 citations
- Reflexion: language agents with verbal reinforcement learningNoah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan et al.NeurIPS 2023 · 5,828 citations
- ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIsYujia Qin, Shihao Liang, Yining Ye, Kunlun Zhu et al.ICLR 2024 · 1,469 citations
- M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge DistillationJianlyu Chen, Shitao Xiao, Peitian Zhang, Kun Luo et al.ACL Findings 2024 · 762 citations
- ExpeL: LLM Agents Are Experiential LearnersAndrew Zhao, Daniel Huang, Quentin Xu, Matthieu Lin et al.AAAI 2024 · 484 citations
Related papers
- RuAG: Learned-rule-augmented Generation for Large Language ModelsYudi Zhang, Pei Xiao, Lu Wang, Chaoyun Zhang et al.ICLR 2025
- MCTS-RAG: Enhancing Retrieval-Augmented Generation with Monte Carlo Tree SearchYunhai Hu, Yilun Zhao, Chen Zhao, Arman CohanEMNLP Findings 2025 · 5 citations
- InstructRAG: Leveraging Retrieval-Augmented Generation on Instruction Graphs for LLM-Based Task PlanningZheng Wang, Shu Xian Teo, Jun Jie Chew, Wei ShiSIGIR 2025 · 4 citations
- Progressive Multimodal Reasoning via Active RetrievalGuanting Dong, Chenghao Zhang, Mengjie Deng, Yutao Zhu et al.ACL 2025
- AirRAG: Autonomous Strategic Planning and Reasoning Steer Retrieval Augmented GenerationWenfeng Feng, Chuzhan Hao, Yuewei Zhang, Guochao Jiang et al.EMNLP Findings 2025 · 3 citations
