Metraipolis › archive › dc_d85bac730bce

MemCodex: Self-Programming Hierarchical Memory for Language Agents — arXiv:2609.39765

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.39765
License (as archived)
CC0-1.0
Archived
2026-10-07T13:13:58.748Z
SHA-256 of the archived text
310342dd9c83c4c19f304379ff8a501dde4d0bec780d078641ec9a508a4714e1
Stable URL of this page
https://metraipolis.com/documento/dc_d85bac730bce

How to cite

MemCodex: Self-Programming Hierarchical Memory for Language Agents — arXiv:2609.39765 — arXiv (en), https://arxiv.org/abs/2609.39765 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d85bac730bce (Metraipolis archive dc_d85bac730bce, sha256 310342dd9c83c4c19f304379ff8a501dde4d0bec780d078641ec9a508a4714e1)
BibTeX
@misc{dcd85bac730bce,
  title = {MemCodex: Self-Programming Hierarchical Memory for Language Agents — arXiv:2609.39765},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.39765},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d85bac730bce; sha256 310342dd9c83c4c19f304379ff8a501dde4d0bec780d078641ec9a508a4714e1},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39765. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agent memory faces heterogeneous access needs: a single-hop question may require one piece of evidence, whereas a multi-hop question must combine evidence from multiple sources. Predefined memory workflows cannot adapt to these varying needs. Recent adaptive methods search or learn over memory components and their compositions, but the design space itself remains predefined. We introduce MemCodex, a self-evolving hierarchical memory system that organizes experience into executable memory programs for summaries, relational knowledge, reusable skills, and latent memory. Open-ended program evolution searches the open design space of layer programs by rewriting how each layer is constructed, indexed, retrieved, and routed, thereby adapting both within-layer implementations and cross-layer composition. At query time, reads traverse the hierarchy from coarse to fine and stop once sufficient evidence is found, descending to the original history when needed. We further develop MemArena, a unified runtime that places heterogeneous data and memory systems behind a common interface. MemCodex improves average task success by 10.1% relative to the strongest adaptive-memory baseline, while using 3.4x fewer context tokens and achieving 2.1x faster inference.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.39765 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T12:44:51.858Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).