M$^\star$: Every Task Deserves Its Own Memory Harness

TL;DR

M$^igstar$ introduces an automated method to evolve task-specific memory systems for language model agents, improving performance across diverse domains by customizing memory design.

Contribution

The paper presents a novel approach that automatically discovers optimized memory architectures through executable program evolution, tailored to each task.

Findings

M$^igstar$ outperforms fixed-memory baselines across multiple benchmarks.

Evolved memory programs show distinct structures for different tasks.

Task-specific memory design enhances agent performance and flexibility.

Abstract

Large language model agents rely on specialized memory systems to accumulate and reuse knowledge during extended interactions. Recent architectures typically adopt a fixed memory design tailored to specific domains, such as semantic retrieval for conversations or skills reused for coding. However, a memory system optimized for one purpose frequently fails to transfer to others. To address this limitation, we introduce M$^\star$, a method that automatically discovers task-optimized memory harnesses through executable program evolution. Specifically, M$^\star$ models an agent memory system as a memory program written in Python. This program encapsulates the data Schema, the storage Logic, and the agent workflow Instructions. We optimize these components jointly using a reflective code evolution method; this approach employs a population-based search strategy and analyzes evaluation failures to iteratively refine the candidate programs. We evaluate M$^\star$ on four distinct benchmarks spanning conversation, embodied planning, and expert reasoning. Our results demonstrate that M$^\star$ improves performance over existing fixed-memory baselines robustly across all evaluated tasks. Furthermore, the evolved memory programs exhibit structurally distinct processing mechanisms for each domain. This finding indicates that specializing the memory mechanism for a given task explores a broad design space and provides a superior solution compared to general-purpose memory paradigms.