Thank you! RAG mainly helps retrieve relevant information, while our approach focuses more on what should be remembered, reinforced, or forgotten over time. I think combining RAG with Ebbinghaus-based memory decay could make long-running multi-agent systems much more efficient and human-like.
