Microsoft Research presents Memora, a scalable memory system for long-horizon AI agents. It separates rich stored memories from lightweight abstractions and cue anchors used for retrieval, aiming to preserve specific details without making every query process the full history. According to Microsoft, Memora establishes new state-of-the-art results on LoCoMo and LongMemEval, outperforming Mem0, RAG, and full-context inference while using up to 98% fewer context tokens. The supplied article says the paper appears at ICML 2026, and Microsoft has released the implementation on GitHub.
No heat snapshots are available in the last 24 hours.