Metraipolis › archive › dc_41ab63fb76fc

What Is This Memory Useful For? Measuring Task-Conditioned Utility in Long-Term Agent Memory — arXiv:2611.03930

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.03930
License (as archived): CC0-1.0
Archived: 2026-11-04T18:33:44.758Z
SHA-256 of the archived text: 9a2834e01849afd3672d1e8f51fb0708c1c74c019924019fcaa63e66e65865d5
Stable URL of this page: https://metraipolis.com/documento/dc_41ab63fb76fc

How to cite

What Is This Memory Useful For? Measuring Task-Conditioned Utility in Long-Term Agent Memory — arXiv:2611.03930 — arXiv (en), https://arxiv.org/abs/2611.03930 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_41ab63fb76fc (Metraipolis archive dc_41ab63fb76fc, sha256 9a2834e01849afd3672d1e8f51fb0708c1c74c019924019fcaa63e66e65865d5)

BibTeX

@misc{dc41ab63fb76fc,
  title = {What Is This Memory Useful For? Measuring Task-Conditioned Utility in Long-Term Agent Memory — arXiv:2611.03930},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.03930},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_41ab63fb76fc; sha256 9a2834e01849afd3672d1e8f51fb0708c1c74c019924019fcaa63e66e65865d5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.03930. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A stored experience may answer one future question and offer little help with another. We measure this task-conditioned utility rather than assigning each memory a single importance score. A task-blind writer stores a memory without seeing the later task; a frozen reader then answers that task with and without the memory, using identical filler. A cost ledger records the resources used to form, store and expose the memory. On single-call factual-recall tasks built from LongMemEval-V2 web-agent memories, the same memory helps more when it contains information needed for the answer. Its mean paired reward gain on these answer-bearing task pairs exceeds that on matched-control pairs by $+0.33$ (95% interval $+0.21$ to $+0.46$). The tested task-blind summaries and structured records yield smaller gains than Raw, including at fixed total context; a budget sweep is consistent with retention of answer-bearing content contributing to the difference. The bearing--control contrast remains positive in a new session with randomized memory position and under a second reader, although individual memory--task labels change. Generic retrieval performs above chance but misses most screen-defined answer-bearing memories. The factual-recall gains do not extend to the tested software-engineering and scientific-coding benchmarks, where even raw memories have no measurable effect. We provide a public, versioned Conditional Memory Utility Dataset of paired measurements and costs. The practical takeaway is to evaluate memory compression by the task utility it preserves alongside its costs. Task-conditioned measurements make this comparison possible; their stability must also be checked.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2611.03930 — 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 2026-11-04T18:32:37.252Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).