Mnemis: Dual-Route Retrieval on Hierarchical Graphs for Long-Term LLM Memory

2026-02-16

AI Summary

This work proposes Mnemis, a novel memory architecture for large language models (LLMs) that integrates dual-process cognition to overcome the limitations of conventional similarity-based retrieval, which often fails to support tasks requiring global reasoning or comprehensive information coverage. Mnemis uniquely combines System-1 fast retrieval—based on semantic similarity—with System-2 deliberate selection grounded in a hierarchical graph structure, thereby establishing a dual-path retrieval mechanism that leverages both semantic and structural relevance. By organizing memories into a graph that enables top-down hierarchical traversal alongside similarity search, Mnemis achieves state-of-the-art performance, scoring 93.9 on LoCoMo and 91.6 on LongMemEval-S using GPT-4.1-mini.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

AI Memory, specifically how models organizes and retrieves historical messages, becomes increasingly valuable to Large Language Models (LLMs), yet existing methods (RAG and Graph-RAG) primarily retrieve memory through similarity-based mechanisms. While efficient, such System-1-style retrieval struggles with scenarios that require global reasoning or comprehensive coverage of all relevant information. In this work, We propose Mnemis, a novel memory framework that integrates System-1 similarity search with a complementary System-2 mechanism, termed Global Selection. Mnemis organizes memory into a base graph for similarity retrieval and a hierarchical graph that enables top-down, deliberate traversal over semantic hierarchies. By combining the complementary strength from both retrieval routes, Mnemis retrieves memory items that are both semantically and structurally relevant. Mnemis achieves state-of-the-art performance across all compared methods on long-term memory benchmarks, scoring 93.9 on LoCoMo and 91.6 on LongMemEval-S using GPT-4.1-mini.

Problem

• AI Memory
• Long-Term Memory
• Similarity-Based Retrieval
• Global Reasoning
• LLM

Innovation

• Dual-Route Retrieval
• Hierarchical Graph
• System-2 Reasoning
• Long-Term LLM Memory
• Global Selection