Search papers, labs, and topics across Lattice.
APEX-MEM, a novel conversational memory system, structures conversations as temporally grounded events within a property graph using a domain-agnostic ontology. This system employs append-only storage to maintain the full temporal evolution of information and uses a multi-tool retrieval agent to resolve conflicting information at query time. Experiments on LOCOMO and LongMemEval show APEX-MEM achieves state-of-the-art performance in question answering and long-term memory tasks, demonstrating the benefits of structured property graphs for temporally coherent reasoning.
Forget simply scaling context windows: structuring conversational memory as a temporally-grounded property graph unlocks significantly more accurate long-term reasoning.
Large language models still struggle with reliable long-term conversational memory: simply enlarging context windows or applying naive retrieval often introduces noise and destabilizes responses. We present APEX-MEM, a conversational memory system that combines three key innovations: (1) a property graph which uses domain-agnostic ontology to structure conversations as temporally grounded events in an entity-centric framework, (2) append-only storage that preserves the full temporal evolution of information, and (3) a multi-tool retrieval agent that understands and resolves conflicting or evolving information at query time, producing a compact and contextually relevant memory summary. This retrieval-time resolution preserves the full interaction history while suppressing irrelevant details. APEX-MEM achieves 88.88% accuracy on LOCOMO's Question Answering task and 86.2% on LongMemEval, outperforming state-of-the-art session-aware approaches and demonstrating that structured property graphs enable more temporally coherent long-term conversational reasoning.
