Metraipolis › archive › dc_c09e32ee5df9
Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classification — arXiv:2608.17247
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.17247
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T14:28:32.901Z
- SHA-256 of the archived text
4144d1cc79f842d3ab2eb92f451a6feef259c753c0c3d9579217a8560422c982- Stable URL of this page
https://metraipolis.com/documento/dc_c09e32ee5df9
How to cite
Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classification — arXiv:2608.17247 — arXiv (en), https://arxiv.org/abs/2608.17247 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c09e32ee5df9 (Metraipolis archive dc_c09e32ee5df9, sha256 4144d1cc79f842d3ab2eb92f451a6feef259c753c0c3d9579217a8560422c982)
BibTeX
@misc{dcc09e32ee5df9,
title = {Explicit State Elicitation Is Not Enough: A Controlled Audit of Memory-Policy Classification — arXiv:2608.17247},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.17247},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c09e32ee5df9; sha256 4144d1cc79f842d3ab2eb92f451a6feef259c753c0c3d9579217a8560422c982},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.17247. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Personalized agents must decide whether retrieved user memory should be used, ignored, updated, or queried before it affects a current task. We use this setting to develop an empirical audit protocol for structured intermediate outputs: first audit dataset shortcuts, then isolate bundled prompt changes, check whether intermediate labels are answer-associated, test decomposed semantic evidence, and audit provider-level execution failures. A 480-example synthetic development set initially suggested large gains from a state-structured prompt bundle, but TF-IDF diagnostics showed lexical separability and no positive standalone Ignore cases. We therefore construct a frozen 160-example controlled counterfactual set with 40 matched four-way families and rule-derived reference policies. On this set, exposing the four state definitions improves accuracy, but an isolated explicit state-output field does not significantly improve policy accuracy for Llama-3.3-70B and gives only a marginal, non-significant gain for GPT-OSS-120B. Supplying benchmark-associated state labels shifts policy predictions, but because those labels deterministically map to policies, this is a label-conditioning diagnostic rather than evidence of a faithful internal mechanism. Family-level and seed-stability analyses further show that example-level accuracy overstates counterfactual consistency: complete four-way family success is rare. An exploratory follow-up that elicits decomposed semantic evidence also fails to improve routing for the cleanly evaluated endpoint; the corresponding GPT-OSS condition was unavailable because of provider-side request validation. We evaluate policy classification only, not downstream responses, tool actions, or memory-store mutation. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-19 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.17247 — 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 2026-10-18T14:27:54.245Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-19?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).