What question did this study set out to answer?

The aim is to compare various memory prioritization algorithms to improve retrieval accuracy in multi-agent systems.

April 10, 2026Open Access

From Static Scores to Cognitive Retrieval: A Comparative Evaluation of Memory Prioritization Algorithms for LLM Multi-Agent Systems

Key Points

The aim is to compare various memory prioritization algorithms to improve retrieval accuracy in multi-agent systems.
Deployment of DRM v1 with five LLM agents managing over 200 memory entries.
Evaluation of seven retrieval algorithms from information retrieval and cognitive science.
Implementation of a three-layer cache architecture leveraging ACT-R dynamics.
Score inflation led to 54% of entries clustering at high scores, reducing retrieval accuracy.
Achieved 100% accuracy across all evaluation scenarios at 2.6 ms per ranking.
Utilized keyword matching without embeddings or vector databases.

Abstract

Paper 2-1 in the Buddys Architecture Series. A supplementary paper to Paper 2 (DRM, DOI: 10.5281/zenodo.19211620). After three weeks of production deployment of DRM v1 with five LLM agents and over 200 memory entries, we observed score inflation: 54% of entries clustered at scores 9-10, degrading top-5 retrieval accuracy to 34.2%. We evaluated seven retrieval algorithms spanning information retrieval, cognitive science (ACT-R), and control theory (DNFO). The selected architecture — a three-layer cache combining category-level softmax gating, pointer-level multi-dimensional scoring with ACT-R activation dynamics, and on-demand content retrieval — achieved 100% accuracy across all six evaluation scenarios at 2.6 ms per ranking for N=200, using only keyword matching over Markdown files. No embeddings, no vector database, no GPU.

Read Full Paperexternally

Mark Helpful

Bookmark

Relay

View Full Paper