Harnessing Long-Term Memory for Adaptive AI Agents

Published in Under Review, 2025

This work explores a memory-augmented agent architecture that integrates long-term episodic and semantic memory mechanisms. The proposed framework enables AI agents to adaptively retrieve, update, and utilize past experiences for improved decision-making and continual learning in dynamic environments.