Metraipolis › archive › dc_7fe1a1e5330e

GenMem: Generative Symbolic Memory for Self-Evolving Harness — arXiv:2609.34633

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.34633
License (as archived)
CC0-1.0
Archived
2026-10-07T22:38:30.746Z
SHA-256 of the archived text
e62fae4f3c7979ac0c038c84d7e624d545ff18b7bfd5f74cfc42689e0f9e97d5
Stable URL of this page
https://metraipolis.com/documento/dc_7fe1a1e5330e

How to cite

GenMem: Generative Symbolic Memory for Self-Evolving Harness — arXiv:2609.34633 — arXiv (en), https://arxiv.org/abs/2609.34633 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7fe1a1e5330e (Metraipolis archive dc_7fe1a1e5330e, sha256 e62fae4f3c7979ac0c038c84d7e624d545ff18b7bfd5f74cfc42689e0f9e97d5)
BibTeX
@misc{dc7fe1a1e5330e,
  title = {GenMem: Generative Symbolic Memory for Self-Evolving Harness — arXiv:2609.34633},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.34633},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7fe1a1e5330e; sha256 e62fae4f3c7979ac0c038c84d7e624d545ff18b7bfd5f74cfc42689e0f9e97d5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.34633. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-term memory supports the self-evolution of LLM agents by retaining experience and skills across tasks and enabling their retrieval, reuse, and revision in subsequent long-horizon decision-making. Yet existing memory management approaches remain limited to discriminative retrieval and to address the sparse, hierarchical, and highly redundant structure of reusable experience: only a small, task-dependent subset of trajectories and memories warrants retention, retrieval, or revision. Learning these operations is further complicated by sparse, delayed, and indirect task-level feedback, with weak supervision across the memory lifecycle. Moreover, continual memory evolution introduces an architectural tension as addressing invariance: stored experience is perpetually revised, yet the addressing interface consumed by learned retrieval policies must remain stable. To address, we present GenMem, which reformulates memory management as generative symbolic addressing. Its core mechanism is the Symbolic Identifier (SID), a multi-level discrete token tuple drawn from a Cartesian-product address space that factorizes a million-scale sparse memory space using fewer than one hundred discrete symbols. Instead of generating ever-changing raw content, the memory agent learns to generate SIDs, while memory evolution rewrites the payload at a fixed address without shifting the address itself. Architecturally, GenMem couples a MemRetriever and a MemEvolver within a multi-agent harness, trained via GRPO with dense process and outcome rewards with two-channels optimization. Under offline memory evolution, experiments spanning ALFWorld, WebShop, multi-hop QA, medical reasoning, and deep research evaluate GenMem against strong memory-augmented baselines...

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.34633 — 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 2026-10-07T22:14:28.257Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).