MMAG: Mixed Memory-Augmented Generation for Large Language Models Applications

TL;DR

This paper presents MMAG, a multi-layered memory framework for LLMs inspired by human cognition, enhancing long-term relevance, personalization, and continuity in conversational agents.

Contribution

Introduces the MMAG framework organizing multiple memory types for LLMs, with implementation in the Heero agent demonstrating improved engagement and retention.

Findings

Encrypted long-term memory improves user engagement.

Multi-layered memory enhances conversational coherence.

Framework addresses privacy, storage, and latency challenges.

Abstract

Large Language Models (LLMs) excel at generating coherent text within a single prompt but fall short in sustaining relevance, personalization, and continuity across extended interactions. Human communication, however, relies on multiple forms of memory, from recalling past conversations to adapting to personal traits and situational context. This paper introduces the Mixed Memory-Augmented Generation (MMAG) pattern, a framework that organizes memory for LLM-based agents into five interacting layers: conversational, long-term user, episodic and event-linked, sensory and context-aware, and short-term working memory. Drawing inspiration from cognitive psychology, we map these layers to technical components and outline strategies for coordination, prioritization, and conflict resolution. We demonstrate the approach through its implementation in the Heero conversational agent, where encrypted long-term bios and conversational history already improve engagement and retention. We further discuss implementation concerns around storage, retrieval, privacy, and latency, and highlight open challenges. MMAG provides a foundation for building memory-rich language agents that are more coherent, proactive, and aligned with human needs.