Multi-Layered Memory Architectures for LLM Agents: An Experimental Evaluation of Long-Term Context Retention
This addresses memory issues in long-term dialogue agents, offering incremental improvements for enhanced retention and reasoning stability.
The paper tackles the problem of semantic drift and unstable memory retention in long-horizon dialogue systems by introducing a Multi-Layer Memory Framework, achieving improved performance with metrics such as 46.85 Success Rate, 0.618 overall F1, and 56.90% six-period retention while reducing false memory rate to 5.1%.
Long-horizon dialogue systems suffer from semanticdrift and unstable memory retention across extended sessions. This paper presents a Multi-Layer Memory Framework that decomposes dialogue history into working, episodic, and semantic layers with adaptive retrieval gating and retention regularization. The architecture controls cross-session drift while maintaining bounded context growth and computational efficiency. Experiments on LOCOMO, LOCCO, and LoCoMo show improved performance, achieving 46.85 Success Rate, 0.618 overall F1 with 0.594 multi-hop F1, and 56.90% six-period retention while reducing false memory rate to 5.1% and context usage to 58.40%. Results confirm enhanced long-term retention and reasoning stability under constrained context budgets.