Empirical-MCTS: Continuous Agent Evolution via Dual-Experience Monte Carlo Tree Search

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lu, Hao, Huang, Haoyuan, Zhou, Yulin, Li, Chen, Zhu, Ningxin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912875366318080
author Lu, Hao
Huang, Haoyuan
Zhou, Yulin
Li, Chen
Zhu, Ningxin
author_facet Lu, Hao
Huang, Haoyuan
Zhou, Yulin
Li, Chen
Zhu, Ningxin
contents Inference-time scaling strategies, particularly Monte Carlo Tree Search (MCTS), have significantly enhanced the reasoning capabilities of Large Language Models (LLMs). However, current approaches remain predominantly stateless, discarding successful reasoning patterns after each problem instance and failing to mimic the empirical accumulation of wisdom characteristic of human problem-solving. To bridge this gap, we introduce Empirical-MCTS, a dual-loop framework that transforms stateless search into a continuous, non-parametric learning process. The framework unifies local exploration with global memory optimization through two novel mechanisms: Pairwise-Experience-Evolutionary Meta-Prompting (PE-EMP) and a Memory Optimization Agent. PE-EMP functions as a reflexive optimizer within the local search, utilizing pairwise feedback to dynamically synthesize adaptive criteria and evolve meta-prompts (system prompts) in real-time. Simultaneously, the Memory Optimization Agent manages a global repository as a dynamic policy prior, employing atomic operations to distill high-quality insights across problems. Extensive evaluations on complex reasoning benchmarks, including AIME25, ARC-AGI-2, and MathArena Apex, demonstrate that Empirical-MCTS significantly outperforms both stateless MCTS strategies and standalone experience-driven agents. These results underscore the critical necessity of coupling structured search with empirical accumulation for mastering complex, open-ended reasoning tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2602_04248
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Empirical-MCTS: Continuous Agent Evolution via Dual-Experience Monte Carlo Tree Search
Lu, Hao
Huang, Haoyuan
Zhou, Yulin
Li, Chen
Zhu, Ningxin
Artificial Intelligence
Computation and Language
Inference-time scaling strategies, particularly Monte Carlo Tree Search (MCTS), have significantly enhanced the reasoning capabilities of Large Language Models (LLMs). However, current approaches remain predominantly stateless, discarding successful reasoning patterns after each problem instance and failing to mimic the empirical accumulation of wisdom characteristic of human problem-solving. To bridge this gap, we introduce Empirical-MCTS, a dual-loop framework that transforms stateless search into a continuous, non-parametric learning process. The framework unifies local exploration with global memory optimization through two novel mechanisms: Pairwise-Experience-Evolutionary Meta-Prompting (PE-EMP) and a Memory Optimization Agent. PE-EMP functions as a reflexive optimizer within the local search, utilizing pairwise feedback to dynamically synthesize adaptive criteria and evolve meta-prompts (system prompts) in real-time. Simultaneously, the Memory Optimization Agent manages a global repository as a dynamic policy prior, employing atomic operations to distill high-quality insights across problems. Extensive evaluations on complex reasoning benchmarks, including AIME25, ARC-AGI-2, and MathArena Apex, demonstrate that Empirical-MCTS significantly outperforms both stateless MCTS strategies and standalone experience-driven agents. These results underscore the critical necessity of coupling structured search with empirical accumulation for mastering complex, open-ended reasoning tasks.
title Empirical-MCTS: Continuous Agent Evolution via Dual-Experience Monte Carlo Tree Search
topic Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2602.04248