AtomMem : Learnable Dynamic Agentic Memory with Atomic Memory Operation
This addresses the need for more flexible and generalizable memory designs in AI agents, though it appears incremental as it builds on existing memory concepts with a learning-based approach.
The paper tackled the problem of inflexible, hand-crafted memory mechanisms in agents for long-horizon tasks by proposing AtomMem, a learnable framework that reframes memory management as dynamic decision-making with atomic CRUD operations, resulting in consistent outperformance over prior static methods across three benchmarks.
Equipping agents with memory is essential for solving real-world long-horizon problems. However, most existing agent memory mechanisms rely on static and hand-crafted workflows. This limits the performance and generalization ability of these memory designs, which highlights the need for a more flexible, learning-based memory framework. In this paper, we propose AtomMem, which reframes memory management as a dynamic decision-making problem. We deconstruct high-level memory processes into fundamental atomic CRUD (Create, Read, Update, Delete) operations, transforming the memory workflow into a learnable decision process. By combining supervised fine-tuning with reinforcement learning, AtomMem learns an autonomous, task-aligned policy to orchestrate memory behaviors tailored to specific task demands. Experimental results across 3 long-context benchmarks demonstrate that the trained AtomMem-8B consistently outperforms prior static-workflow memory methods. Further analysis of training dynamics shows that our learning-based formulation enables the agent to discover structured, task-aligned memory management strategies, highlighting a key advantage over predefined routines.