Metraipolis › archive › dc_ec64ccbd8282

KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents — arXiv:2610.41482

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.41482
License (as archived): CC0-1.0
Archived: 2026-11-02T07:01:10.864Z
SHA-256 of the archived text: e14175a4e6114105c746c9d7314972838f09a1a8935f9ad47ec57df56f257a9e
Stable URL of this page: https://metraipolis.com/documento/dc_ec64ccbd8282

How to cite

KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents — arXiv:2610.41482 — arXiv (en), https://arxiv.org/abs/2610.41482 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ec64ccbd8282 (Metraipolis archive dc_ec64ccbd8282, sha256 e14175a4e6114105c746c9d7314972838f09a1a8935f9ad47ec57df56f257a9e)

BibTeX

@misc{dcec64ccbd8282,
  title = {KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents — arXiv:2610.41482},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.41482},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ec64ccbd8282; sha256 e14175a4e6114105c746c9d7314972838f09a1a8935f9ad47ec57df56f257a9e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41482. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recursive textual memory can carry earlier errors into later updates, yet task performance alone does not establish memory reliability. We define self-consistency bias as the persistence of incorrect claims already present in the incoming memory. We present KEEPER, a comparative analysis of four auxiliary objectives along information-content and update-sensitivity axes within a shared recursive interface, examining how supervision targets relate to agent performance. Among evaluated conditions, reward-related prediction achieve the highest task performance before and after offline reinforcement learning, outperforming joint supervision. A claim-level analysis of one baseline trajectory documents error persistence despite corrective observations. Together, these analyses motivate evaluating error persistence and valid-information retention alongside task outcomes.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.41482 — 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 2026-11-02T06:56:58.397Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).