Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

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

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

Abstract

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.

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

Abstract

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.

Paper Structure

This paper contains 38 sections, 13 equations, 1 figure, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. LLM Agents.
  4. Neuro-Symbolic Agents.
  5. Method
  6. Overview
  7. Feasibility Alignment via Symbolic Memory
  8. Progress Alignment via Neural Memory
  9. Dual-Alignment Inference via Neuro-Symbolic Memory
  10. Experiments
  11. Experimental Setup
  12. Competing Baselines.
  13. Ablation Setting.
  14. Main Results
  15. Ablation Studies
  16. ...and 23 more sections

Figures (1)

  • Figure 1: Overview of our neuro-symbolic dual memory framework. The proposed system explicitly separates local feasibility alignment from global progress alignment according to the distinct reasoning demands of the two objectives. Top (Offline Phase): Failed interactions are compiled into executable symbolic verifier rules to construct the symbolic Feasibility Memory, while successful trajectories are distilled into stage-anchored procedural blueprints to form the neural Progress Memory. Bottom (Inference Phase): At inference time, Progress Memory provides stage-aware guidance for proposing progress-consistent actions, while Feasibility Memory performs symbolic feasibility verification and iterative refinement before execution.