Metraipolis › archive › dc_b3267d07214a

Memory That Changes Action Is Not Memory That Guides It: Counterfactual Auditing of History-Conditioned Robot Policies — arXiv:2609.27247

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.27247
License (as archived)
CC0-1.0
Archived
2026-10-08T12:53:27.402Z
SHA-256 of the archived text
eea86b3bbab630c105d92820adccc037f9cdb3a9082e8d5e41ec9e96560e8170
Stable URL of this page
https://metraipolis.com/documento/dc_b3267d07214a

How to cite

Memory That Changes Action Is Not Memory That Guides It: Counterfactual Auditing of History-Conditioned Robot Policies — arXiv:2609.27247 — arXiv (en), https://arxiv.org/abs/2609.27247 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b3267d07214a (Metraipolis archive dc_b3267d07214a, sha256 eea86b3bbab630c105d92820adccc037f9cdb3a9082e8d5e41ec9e96560e8170)
BibTeX
@misc{dcb3267d07214a,
  title = {Memory That Changes Action Is Not Memory That Guides It: Counterfactual Auditing of History-Conditioned Robot Policies — arXiv:2609.27247},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.27247},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b3267d07214a; sha256 eea86b3bbab630c105d92820adccc037f9cdb3a9082e8d5e41ec9e96560e8170},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27247. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A robot returning a block to its origin tray may encounter two task-consistent pasts that reconverge to the same current input but warrant different actions. Yet memory-policy evaluations often rely on task success or action change under memory perturbation, neither of which establishes that memory guides the decision. We propose the \textbf{Counterfactual Memory Audit (CMA)}, an evaluation protocol that crosses two histories at a verified-identical present, queries a frozen policy under common randomness, and evaluates each saved action under both pasts. This separates memory sensitivity, warranted choice, matched-world physical value, and per-pair reliability. On Mem-0, every audited Put Back pair changes action, but only $20/64$ pairs are fully reliable; at a later Swap decision, all paired actions change while both memories select the same branch. Native interventions further show closed-loop influence: replacing the history bank redirects behavior toward the replaced content, while restoring a 4096-byte protected anchor recovers $38.9$ points of Swap success lost to injected bank faults. On a dual-arm physical platform, memory changes saved actions, yet five of nine completed Put Back manipulations reach the wrong target. These results show that a robot can remember and react without reliably using memory to choose the behavior its past warrants. CMA provides a decision-level audit for distinguishing these cases.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-09-24 in cs.RO).
Categorie arXiv: Robotics (cs.RO).
Pagina del lavoro: https://arxiv.org/abs/2609.27247 — 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 2026-10-08T12:21:19.448Z dalla pagina https://arxiv.org/catchup/cs.RO/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).