EverMemOS

Self-organizing memory OS for structured long-horizon reasoning. Three-phase model: episodic, semantic, reconstructive.

At a glance

Optimised for: research positioning (see memory_model)

Anti-fit: not applicable - research paper

Memory-OS framework targeting always-on agents.

Preprint; production deployment evidence light.

Reports 83.0% overall accuracy on LongMemEval.

Other research / specialised systems in the catalog, ranked by inbound references.

A-MEM T3
Treats memories as atomic linkable notes — explicit nod to Zettelkasten knowledge management. Dynamic linking; retroactive memory revision.
BAI-LAB MemoryOS T3
Hierarchical "OS" with Storage / Updating / Retrieval / Generation modules. Short-term → mid-term via FIFO dialogue-chain; mid-term → long-term via segmented paging.
Titans (Google) T4
Neural long-term memory module that learns to memorise at test time. Uses gradient-of-loss as "surprise" signal; adaptive weight-decay forgetting. Three variants: MAC (memory-as-context), MAG (memory-as-gate), MAL (memory-as-layer).
EVOLVE-MEM T3
Dynamic Memory Network + Hierarchical Memory Manager + Self-Improvement Engine. L0 raw embeddings, L1 contextual summaries, L2 high-level principles. NeurIPS 2025 (Scaling Environments for Agents workshop).
LiCoMemory T4
Lightweight hierarchical graph (CogniGraph) with entities and relations as semantic indexing layers. Incremental graph construction, fast updates, low-latency inference. Nov 2025.
MemoRAG T3
RAG framework on top of a long-context memory model. Builds global memory once, generates contextual clues at query time. TheWebConf 2025.