Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity
Microsoft Research Xuchao Zhang, Molly Xia, Mayukh Das, Anson Bastos, Rujia Wang, Chetan Bansal, Saravan Rajmohan
Microsoft researchers developed Memora, a memory system for AI agents that separates what is stored (rich content) from how it is retrieved (lightweight abstractions and cue anchors) to balance detail preservation with efficient scaling. Memora achieved state-of-the-art performance on LoCoMo and LongMemEval benchmarks while reducing token consumption by up to 98% compared to full-context inference and halving the memory entries needed versus Mem0. The system enables AI agents to maintain detailed project histories and context over long-horizon tasks without repeatedly re-reading entire conversation histories.
Why it matters
AI agents can't remember past conversations. They must constantly reload or retrieve context, which grows less efficient as tasks get longer and more complex. Memora solves this with a scalable memory system separating what’s stored from how it's retrieved. The post Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity appeared first on Microsoft Research.
