Structurally Aligned Subtask-Level Memory for Software Engineering Agents

TL;DR

This paper introduces a subtask-level memory mechanism for LLM-based software engineering agents, improving long-horizon reasoning by aligning memory with the agent's functional decomposition, leading to better performance on complex tasks.

Contribution

It proposes Structurally Aligned Subtask-Level Memory, addressing granularity mismatch in prior instance-level memory approaches, and demonstrates its effectiveness in SWE tasks.

Findings

Outperforms vanilla agents and instance-level memory baselines

Improves mean Pass@1 by +4.7 percentage points on average

Gains increase with more interaction steps

Abstract

Large Language Models (LLMs) have demonstrated significant potential as autonomous software engineering (SWE) agents. Recent work has further explored augmenting these agents with memory mechanisms to support long-horizon reasoning. However, these approaches typically operate at a coarse instance granularity, treating the entire problem-solving episode as the atomic unit of storage and retrieval. We empirically demonstrate that instance-level memory suffers from a fundamental granularity mismatch, resulting in misguided retrieval when tasks with similar surface descriptions require distinct reasoning logic at specific stages. To address this, we propose Structurally Aligned Subtask-Level Memory, a method that aligns memory storage, retrieval, and updating with the agent's functional decomposition. Extensive experiments on SWE-bench Verified demonstrate that our method consistently outperforms both vanilla agents and strong instance-level memory baselines across diverse backbones, improving mean Pass@1 over the vanilla agent by +4.7 pp on average (e.g., +6.8 pp on Gemini 2.5 Pro). Performance gains grow with more interaction steps, showing that leveraging past experience benefits long-horizon reasoning in complex software engineering tasks.