Home▸AI & Machine Learning▸2D Agent Memory Retrieval: Recency Decay, Cosine Relevance, Importance Weights and Top-k Recall

🧠 2D Agent Memory Retrieval: Recency Decay, Cosine Relevance, Importance Weights and Top-k Recall

Score-space view of an agent memory store: recency decay, cosine relevance and importance combined with adjustable weights, top-k selection, precision and recall against a ground truth, verified by a hand cosine value, an Euler decay integral, heap versus sort and a hypergeometric baseline.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted🟠 Blender⇄ 3D version
2D Agent Memory Retrieval: Recency Decay, Cosine Relevance, Importance Weights and Top-k Recall ↗ Open standalone
⚙ Under the hood

Score-space view of an agent memory store: recency decay, cosine relevance and importance combined with adjustable weights, top-k selection, precision and recall against a ground truth, verified by a hand cosine value, an Euler decay integral, heap versus sort and a hypergeometric baseline.

Agent MemoryRetrievalCosine SimilarityRecency DecayImportance ScoreTop-kPrecision and RecallEmbeddings

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)