Sequential Retrieval in Dense Associative Memory: Asymptotic Dynamics and Storage Capacity

@misc{mishima2026sequential,
  title = {{Sequential Retrieval in Dense Associative Memory: Asymptotic Dynamics and Storage Capacity}},
  author = {Mao Mishima and Ayaka Sakata and Kazushi Mimura},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2609.13987},
  eprint = {2609.13987},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.13987},
}

TY  - UNPB
TI  - Sequential Retrieval in Dense Associative Memory: Asymptotic Dynamics and Storage Capacity
AU  - Mao Mishima
AU  - Ayaka Sakata
AU  - Kazushi Mimura
PY  - 2026
DA  - 2026/10/27
PB  - arXiv
DO  - 10.48550/arxiv.2609.13987
UR  - https://arxiv.org/abs/2609.13987
AB  - We study the retrieval dynamics of an n-body asymmetric dense associative memory model, in which sequential retrieval is implemented by shifting one pattern index in the higher-order Hebbian interaction. The model exhibits sequential retrieval of the stored patterns, and the retrieval state corresponds to a stable limit-cycle attractor. The storage capacity is defined as the maximum loading rate for which this limit cycle remains stable. Using generating functional analysis, we derive the dynamical equations describing the sequential retrieval state. The generating functional analysis reveals that the retarded self-interaction term vanishes in the asymmetric model, and the resulting overlap dynamics is reproduced by an Amari-Maginu-type signal-to-noise analysis. Also, we find that for the interaction order n≥ 3, the normalized storage capacity decreases monotonically with increasing interaction order, in contrast to symmetric Modern Hopfield models.
ER  -