arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.02734cs.AI
收藏

对齐进展与可行性:一种神经符号双记忆框架用于长周期LLM智能体

Aligning Progress and Feasibility: A Neuro-Symbolic Dual Memory Framework for Long-Horizon LLM Agents

Nanjing University(南京大学);Jilin University(吉林大学)

Bin Wen, Ruoxuan Zhang, Yang Chen, Hongxia Xie, Lan-Zhe Guo

AI总结:

本文提出神经符号双记忆框架,通过分离语义进展指导与逻辑可行性验证,解决长周期任务中进展漂移与可行性违反问题,实验显示在ALFWorld、WebShop和TextCraft上性能优于现有方法。

摘要:

大型语言模型(LLMs)在长周期决策任务中展现出强大潜力,如具身操作和网络交互。然而,智能体常陷入无限试错循环或偏离主要目标。本文将失败归因于两种根本性错误:全局进展漂移和局部可行性违反。现有方法通常通过单一范式同时解决两者,但二者本质不同:前者依赖模糊语义规划,后者需要严格逻辑约束和状态验证。单一范式方法的固有局限性对处理长周期任务构成根本挑战。受此启发,本文提出神经符号双记忆框架,明确分离语义进展指导与逻辑可行性验证。在推理阶段,框架同步调用两种记忆机制:一方面,基于神经网络的进展记忆从成功轨迹中提取语义蓝图以指导全局任务进展;另一方面,基于符号逻辑的可行性记忆利用从失败转换合成的可执行Python验证函数进行严格逻辑验证。实验表明,该方法在ALFWorld、WebShop和TextCraft上显著优于现有竞争基线,同时大幅减少无效动作率和平均轨迹长度。

英文摘要:

Large language models (LLMs) have demonstrated strong potential in long-horizon decision-making tasks, such as embodied manipulation and web interaction. However, agents frequently struggle with endless trial-and-error loops or deviate from the main objective in complex environments. We attribute these failures to two fundamental errors: global Progress Drift and local Feasibility Violation. Existing methods typically attempt to address both issues simultaneously using a single paradigm. However, these two challenges are fundamentally distinct: the former relies on fuzzy semantic planning, while the latter demands strict logical constraints and state validation. The inherent limitations of such a single-paradigm approach pose a fundamental challenge for existing models in handling long-horizon tasks. Motivated by this insight, we propose a Neuro-Symbolic Dual Memory Framework that explicitly decouples semantic progress guidance from logical feasibility verification. Specifically, during the inference phase, the framework invokes both memory mechanisms synchronously: on one hand, a neural-network-based Progress Memory extracts semantic blueprints from successful trajectories to guide global task advancement; on the other hand, a symbolic-logic-based Feasibility Memory utilizes executable Python verification functions synthesized from failed transitions to perform strict logical validation. Experiments demonstrate that this method significantly outperforms existing competitive baselines on ALFWorld, WebShop, and TextCraft, while drastically reducing the invalid action rate and average trajectory length.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑