Preprint · 2026 · Open access

Deployment-Time Memorization in Foundation-Model Agents

Lei, chen, Guilin Zhang, Kai Zhao, Dalmo Cirne, Andy Olsen, Xu Chu, Zeke Miller, Alet Blanken, Amine Anoun, Jerry Ting

arXiv (Cornell University) · Published 8 June 2026 · DOI 10.48550/arxiv.2606.10062

Open PDFView at the publisherCite this paper

In brief

“Deployment-Time Memorization in Foundation-Model Agents” is a 2026 preprint by Lei and 10 co-authors, published in arXiv (Cornell University). OpenAlex counts 6 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at arXiv (Cornell University).

Citations
6
Most cited in
2026 (6)
References
15
Authors
11

Abstract

Foundation-model agents are increasingly long-lived systems that remember users across interactions, making memorization an explicit deployment-time function rather than solely a property of model weights. Existing work addresses parametric memorization or audits fixed memory configurations, but does not characterize how memory-design choices jointly shape personalization utility, extraction risk, and deletion fidelity. We study this surface as deployment-time memorization, formulating agent memory as a privacy-utility frontier measured by Personalization Recall (PR) and Adversarial Extraction Rate (AER), and sweeping three memory-design knobs: summarization aggressiveness, retrieval breadth (k), and deletion mode. We further introduce the Forgetting Residue Score (FRS) to quantify whether deleted information remains recoverable from derived memory tiers. On LongMemEval, key-fact summarization reduces canary extraction by 76% on Gemma 3 12B and 64% on GPT-4o-mini while preserving nearly all personalization recall; critically, once content is compressed away, increasing k no longer restores leakage. The same compression, however, induces a deletion-fidelity failure: raw-only deletion leaves derived summary copies recoverable in approximately 20% of instances, and only full-pipeline purge or tombstone redaction drives worst-tier residue to zero. Together, these results establish that persistent agent memory must be evaluated as a first-class memorization mechanism -- assessed by what it helps agents recall, what it makes extractable, and what it can truly erase.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Lei, chen, Zhang, G., Zhao, K., Cirne, D., Olsen, A., Chu, X., Miller, Z., Blanken, A., Anoun, A., & Ting, J. (2026). Deployment-Time Memorization in Foundation-Model Agents. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.10062

MLA 9

Lei, et al. “Deployment-Time Memorization in Foundation-Model Agents.” arXiv (Cornell University), 2026. https://doi.org/10.48550/arxiv.2606.10062.

Chicago

Lei, chen, Guilin Zhang, Kai Zhao, Dalmo Cirne, Andy Olsen, Xu Chu, et al. 2026. “Deployment-Time Memorization in Foundation-Model Agents.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.10062.

Harvard

Lei et al. (2026) ‘Deployment-Time Memorization in Foundation-Model Agents’, arXiv (Cornell University). doi:10.48550/arxiv.2606.10062.

IEEE

Lei et al., “Deployment-Time Memorization in Foundation-Model Agents,” arXiv (Cornell University), 2026, doi: 10.48550/arxiv.2606.10062.

BibTeX

@misc{lei2026deployment,
  title     = {{Deployment-Time Memorization in Foundation-Model Agents}},
  author    = {Lei and chen and Zhang, Guilin and Zhao, Kai and Cirne, Dalmo and Olsen, Andy and Chu, Xu and Miller, Zeke and Blanken, Alet and Anoun, Amine and Ting, Jerry},
  howpublished = {arXiv (Cornell University)},
  year      = {2026},
  doi       = {10.48550/arxiv.2606.10062},
  url       = {https://doi.org/10.48550/arxiv.2606.10062}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data

  • memory design
  • privacy-utility trade-off
  • Gemma
  • GPT-4o-mini

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Lei, chen, Zhang, G., Zhao, K., Cirne, D., Olsen, A., Chu, X., Miller, Z., Blanken, A., Anoun, A., & Ting, J. (2026). Deployment-Time Memorization in Foundation-Model Agents. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2606.10062

How many times has this paper been cited?

6 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Lei, chen, Guilin Zhang, Kai Zhao, Dalmo Cirne, Andy Olsen, Xu Chu, Zeke Miller, Alet Blanken, Amine Anoun, Jerry Ting.

What is the DOI of this paper?

10.48550/arxiv.2606.10062, permanently at https://doi.org/10.48550/arxiv.2606.10062. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.