Metraipolis › archive › dc_5e93dd982eb7
Causal Memory Policy: Making Memory Utility Identifiable by Intervening on Retrieval — arXiv:2610.02070
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.02070
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T12:06:57.194Z
- SHA-256 of the archived text
6771981b878f74de37b5879b824e0fcad008493b6bd75a4f0538c6b9b2b01de4- Stable URL of this page
https://metraipolis.com/documento/dc_5e93dd982eb7
How to cite
Causal Memory Policy: Making Memory Utility Identifiable by Intervening on Retrieval — arXiv:2610.02070 — arXiv (en), https://arxiv.org/abs/2610.02070 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5e93dd982eb7 (Metraipolis archive dc_5e93dd982eb7, sha256 6771981b878f74de37b5879b824e0fcad008493b6bd75a4f0538c6b9b2b01de4)
BibTeX
@misc{dc5e93dd982eb7,
title = {Causal Memory Policy: Making Memory Utility Identifiable by Intervening on Retrieval — arXiv:2610.02070},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.02070},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5e93dd982eb7; sha256 6771981b878f74de37b5879b824e0fcad008493b6bd75a4f0538c6b9b2b01de4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.02070. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Memory-augmented large language models must decide which memories to retain, and recent systems do so by estimating each memory's effect on task performance. However, these estimates rely entirely on retrieved memories. When a memory is never retrieved, store-level interventions produce identical outcomes, leaving its utility unidentified. This is a retrieval-level positivity violation, invisible to diagnostics that examine only memory operations. We introduce Causal Memory Policy (CMP), a causal framework that restores identification by intervening on retrieval itself, reserving a fixed number of context slots for memories sampled with known propensities. CMP estimates memory utility by self-normalized inverse propensity weighting under a balanced assignment design. We prove the causal factorization of memory utility through retrieval, the unbiasedness and exact variance of the estimator, and the optimal decision rule under irreversible operations. Empirically, identification fails for 54% of required memories on LongMemEval and 67% on LoCoMo, and the failure persists in a deployed memory system. CMP improves discrimination between required and non-required memories from 0.54 to 0.66 AUC. Finally, we show that identified memory utility alone is insufficient for retention decisions: per-query utility reaches 0.78 AUC on the query for which it is estimated, yet no aggregation available to a retention policy predicts a memory's value on unseen queries. Code is available at: this https URL . Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-02 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.02070 — 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 2026-10-06T10:21:27.998Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).