Preprint · 2026 · Open access
Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models
Zenodo (CERN European Organization for Nuclear Research) · Published 14 October 2026 · DOI 10.5281/zenodo.23357194
In brief
“Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models” is a 2026 preprint by Jincheng Zhang, published in Zenodo (CERN European Organization for Nuclear Research). OpenAlex counts 0 citations as of 19 November 2026. It is open access under a CC BY licence: the full text is free to read at Zenodo (CERN European Organization for Nuclear Research).
- Citations
- 0
- Authors
- 1
Abstract
Large language model (LLM) agents increasingly rely on persistent memory to support long-term reasoning, task continuity, and collaborative problem-solving. However, multi-agent memory systems face a fundamental design trade-off between private memory, which preserves agent-specific information and reduces unnecessary information exchange, and shared memory, which enables knowledge transfer and collective problem-solving. Relying exclusively on private memory may lead to redundant exploration, fragmented knowledge, and repeated errors, whereas unrestricted shared memory may introduce irrelevant information, propagate incorrect conclusions, increase communication costs, and expose sensitive information. This paper proposes Adaptive Private–Shared Memory Switching (APS-MS), a conceptual framework in which multiple LLM agents dynamically select among private memory, shared memory, and hybrid memory according to task requirements and observed system conditions. The framework integrates task-state assessment, memory retrieval, evidence verification, access control, adaptive switching, and feedback-based policy optimization. Rather than treating memory organization as a fixed architectural choice, APS-MS formulates memory selection as a sequential decision-making problem in which the system balances task utility, memory relevance, information reliability, collaboration benefits, latency, computational cost, and privacy risk. The paper presents a formal problem formulation, a reference architecture, an adaptive switching algorithm, a proposed experimental methodology, evaluation metrics, research questions, and testable hypotheses. Because the framework has not yet been experimentally evaluated in this paper, its expected advantages are treated as hypotheses rather than established results. The proposed approach provides a basis for investigating whether adaptive memory coordination can improve the efficiency, reliability, and privacy of multi-LLM systems.
Cite this paper
Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.
APA 7
Zhang, J. (2026). Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models. Zenodo (CERN European Organization for Nuclear Research). https://doi.org/10.5281/zenodo.23357194
MLA 9
Zhang, Jincheng. “Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models.” Zenodo (CERN European Organization for Nuclear Research), 2026. https://doi.org/10.5281/zenodo.23357194.
Chicago
Zhang, Jincheng. 2026. “Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models.” Zenodo (CERN European Organization for Nuclear Research). https://doi.org/10.5281/zenodo.23357194.
Harvard
Zhang, J. (2026) ‘Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models’, Zenodo (CERN European Organization for Nuclear Research). doi:10.5281/zenodo.23357194.
IEEE
J. Zhang, “Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models,” Zenodo (CERN European Organization for Nuclear Research), 2026, doi: 10.5281/zenodo.23357194.
BibTeX
@misc{zhang2026adaptive,
title = {{Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models}},
author = {Zhang, Jincheng},
howpublished = {Zenodo (CERN European Organization for Nuclear Research)},
year = {2026},
doi = {10.5281/zenodo.23357194},
url = {https://doi.org/10.5281/zenodo.23357194}
}
Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)
Topic and keywords
Physical Sciences › Computer Science › Artificial Intelligence › AI-based Problem Solving and Planning
- LLM agents
- multi-agent systems
- persistent memory
- private memory
- shared memory
- hybrid memory
- memory selection
- long-term reasoning
Work with this paper
Most-cited papers in Computer Science
- ImageNet classification with deep convolutional neural networksAlex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton · Communications of the ACM · 2017 · 114,726 citations
- Long Short-Term MemorySepp Hochreiter, Jürgen Schmidhuber · Neural Computation · 1997 · 109,097 citations
- A Mathematical Theory of CommunicationClaude E. Shannon · Bell System Technical Journal · 1948 · 91,736 citations
- Deep learningYann LeCun, Yoshua Bengio, Geoffrey E. Hinton · Nature · 2015 · 90,226 citations
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksShaoqing Ren, Kaiming He, Ross Girshick and 1 more · IEEE Transactions on Pattern Analysis and Machine Intelligence · 2016 · 74,143 citations
Questions about this paper
How do I cite this paper in APA style?
Zhang, J. (2026). Adaptive Switching Between Private Memory and Shared Memory for Multiple Large Language Models. Zenodo (CERN European Organization for Nuclear Research). https://doi.org/10.5281/zenodo.23357194
How many times has this paper been cited?
0 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 Zenodo (CERN European Organization for Nuclear Research): sign in and open the full text from this page.
Who wrote this paper?
Jincheng Zhang.
What is the DOI of this paper?
10.5281/zenodo.23357194, permanently at https://doi.org/10.5281/zenodo.23357194. 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.