Search papers, labs, and topics across Lattice.
The paper introduces Mnemis, a memory framework for LLMs that combines similarity-based retrieval with a novel "Global Selection" mechanism for hierarchical graph traversal. Mnemis addresses the limitations of existing RAG and Graph-RAG methods that struggle with global reasoning by organizing memory into a base graph for similarity search and a hierarchical graph for top-down traversal. Experiments demonstrate that Mnemis achieves state-of-the-art performance on long-term memory benchmarks (LoCoMo and LongMemEval-S) using GPT-4.1-mini.
LLMs can now leverage a hierarchical graph structure for memory retrieval, enabling global reasoning and boosting performance on long-term memory benchmarks beyond what's achievable with similarity search alone.
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.
NVIDIA