{
  "abstract": "Equilibrium Propagation (EqProp) updates each synapse by comparing a local term at the free and nudged equilibria. Because these two values occur at different times, a physical implementation requires temporary local memory. We show that this memory can emerge from a two-timescale synaptic model in which neural and volatile synaptic dynamics are jointly derived from a single energy function. A volatile state $w$ relaxes toward the product of the pre- and postsynaptic activities, while a non-volatile state $M$ stores the learned parameter. At equilibrium, $w$ equals the local term entering the EqProp update. Using its net change between phases implements phase-based EqProp, while transferring its successive changes during nudging yields continual EqProp, in which the temporal comparison is carried by the synaptic dynamics themselves. On MNIST with a single-hidden-layer network, the phase-based model reaches $2.01 \\pm 0.09\\%$ test error versus $2.00 \\pm 0.13\\%$ for standard EqProp, and the continual model reaches $2.24 \\pm 0.11\\%$ versus $2.28 \\pm 0.16\\%$ for continual EqProp. The framework further specifies the dynamical requirements that a volatile--non-volatile synaptic device must satisfy to implement this learning mechanism.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2610.18058",
  "author_total": 2,
  "authors": [
    {
      "name": "Dongshu Liu"
    },
    {
      "name": "Julie Grollier"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/1vtazagtg82za.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.18058",
  "fulltext_url": null,
  "id": "1vtazagtg82za",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.18058",
  "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.ET"
    ],
    [
      "arxiv",
      "physics.app-ph"
    ],
    [
      "openalex",
      "2208|Advanced Memory and Neural Computing"
    ]
  ],
  "openalex": "W7221835838",
  "pdf_url": "https://arxiv.org/pdf/2610.18058",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-13",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/1vtazagtg82za.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:29:11Z",
      "native_id": "2610.18058",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.18058"
    }
  ],
  "subjects": [
    {
      "field": "physics",
      "origin": "arxiv:physics.app-ph",
      "subfield": "applied-instrumentation"
    }
  ],
  "title": "Two-State Dynamical Synapses as a Physical Substrate for Equilibrium Propagation",
  "title_text": "Two-State Dynamical Synapses as a Physical Substrate for Equilibrium Propagation",
  "type": "preprint",
  "updated_at": "2026-10-30T12:29:11Z",
  "url": "https://research.draconitic.com/paper/1vtazagtg82za",
  "venue": "arXiv",
  "versions": []
}