SuperLocalMemory V3: Information-Geometric Foundations for Zero-LLM Enterprise Agent Memory

Summary

Retrieval quality is usually bought with LLM calls. V3 buys it with geometry instead: memories live on a statistical manifold where relevance is a distance, so ranking is computed rather than generated. The result is a benchmark score competitive with cloud systems, produced entirely on-device.

Our summary of the paper. The full abstract is on arXiv. Preprint, not peer reviewed.

Authors

• Varun Pratap Bhardwaj

Cite this paper

@misc{bhardwaj2026slmv3,
  title        = {SuperLocalMemory V3: Information-Geometric Foundations for Zero-LLM Enterprise Agent Memory},
  author       = {Varun Pratap Bhardwaj},
  year         = {2026},
  eprint       = {2603.14588},
  archivePrefix= {arXiv},
  primaryClass = {cs.AI},
  url          = {https://arxiv.org/abs/2603.14588}
}