Search papers, labs, and topics across Lattice.
This paper introduces MindMemOS, a self-evolving memory operating layer designed for AI agents that enhances their ability to adapt and refine memory models over time. By employing the MindMemEvolve algorithm, it utilizes validation-driven evolutionary search to optimize memory schemas, while the dreaming process consolidates memories and resolves conflicts. The system demonstrates impressive performance, achieving 94.03% accuracy on LOCOMO and a 9.2 percentage point improvement on SpreadsheetBench, underscoring its effectiveness in long-term memory adaptation and skill evolution.
MindMemOS enables AI agents to autonomously refine their memories and skills, achieving state-of-the-art accuracy in dynamic environments.
Memory is a core component of AI agents, enabling them to accumulate experience, maintain personalization, and adapt over long-term interactions. However, existing memory systems often remain fixed after development, limiting their ability to adapt their memory models, organization strategies, and procedural knowledge through continued use. We present MindMemOS, a portable and self-evolving memory operating layer that organizes open-world information using a unified entity property timestructure. MindMemOS supports scenario-adaptive memory modeling, higher-order pattern discovery, autonomous memory refinement, and continuous skill evolution. Its MindMemEvolve algorithm employs validation-driven evolutionary search to optimize memory schemas for target scenarios, whiledreaming consolidates accumulated memories by merging redundant records and resolving conflicts. In addition, implicit corrective feedback serves as a human-in-the-loop signal for identifying and revising potentially inaccurate or misaligned memories. Its MindSkillEvolve algorithm further transforms agent execution trajectories into reusable and progressively refined skills. MindMemOS achieves 94.03% accuracy on LOCOMO and 70.63% on PersonaMem. MindSkillEvolve improves SpreadsheetBench success by 9.2 percentage points over the initial-skill baseline.
