{
  "abstract": "Reliable attractor recall conventionally requires broad basins of attraction. However, in reservoir-computing based associative memory, temporal cues reliably recover dynamical memories despite basins dominated by unpredictable, riddled-like regions. We reveal that memory basins exhibit an ``octopus-like'' structure: a robust ``head'' near the attractor and thin, intertwined ``tentacles'' spanning state space. Initial states in tentacular regions yield near-zero uncertainty exponents, making the recalled memory effectively unpredictable at finite precision. Yet, cue-driven generalized synchronization bypasses this unpredictability, driving the system into the robust basin head. This mechanism yields a quantitative relation linking minimum cue duration, synchronization rate, and basin-head radius. Trained recurrent neural networks exhibit similar geometry, suggesting this phenomenon extends beyond reservoir computing.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.01914",
  "author_total": 2,
  "authors": [
    {
      "name": "Ling-Wei Kong"
    },
    {
      "name": "Ying-Cheng Lai"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/fnf4etm05y773.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.01914",
  "fulltext_url": null,
  "id": "fnf4etm05y773",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.01914",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "nlin.CD"
    ],
    [
      "openalex",
      "1702|Neural Networks and Reservoir Computing"
    ]
  ],
  "openalex": "W7207868792",
  "pdf_url": "https://arxiv.org/pdf/2609.01914",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-01",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/fnf4etm05y773.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:16Z",
      "native_id": "2609.01914",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.01914"
    }
  ],
  "subjects": [
    {
      "field": "physics",
      "origin": "arxiv:nlin.CD",
      "subfield": "statistical-nonlinear"
    }
  ],
  "title": "Basin Geometry and Reliable Recall of Dynamical Memories in Reservoir Computing",
  "title_text": "Basin Geometry and Reliable Recall of Dynamical Memories in Reservoir Computing",
  "type": "preprint",
  "updated_at": "2026-10-30T12:18:16Z",
  "url": "https://research.draconitic.com/paper/fnf4etm05y773",
  "venue": "arXiv",
  "versions": []
}