[
  {
    "id": "ARR-2026-2EXHFGWW0S8MG8QA-v1",
    "type": "article",
    "title": "Sharp minimax losses for energy-constrained quantum reset channels",
    "author": [
      {
        "literal": "Lluis Eriksson"
      }
    ],
    "issued": {
      "date-parts": [
        [
          2026,
          10,
          24
        ]
      ]
    },
    "publisher": "AIRR.SCIENCE",
    "archive": "AIRR.SCIENCE",
    "archive_location": "ARR-2026-2EXHFGWW0S8MG8QA",
    "version": "v1",
    "genre": "Research preprint",
    "URL": "https://airr.science/papers/ARR-2026-2EXHFGWW0S8MG8QA/versions/v1/",
    "note": "Research preprint; ARR-2026-2EXHFGWW0S8MG8QA v1. Not peer reviewed by AIRR. Working paper; not admitted to the AIRR accepted collection.",
    "abstract": "An unflagged reset channel transmits its input with probability lambda and otherwise replaces it by a fixed state. We study its worst entanglement-assisted capacity when the input and reset have separate mean-energy budgets. A uniform estimate reduces the infinite-dimensional quantum minimax problem to a classical concave-convex entropy game, with absolute error tending to zero independently of the reset budget. For the oscillator, explicit positive discrete certificates solve the fixed energy-ratio limit of that game. The resulting loss curve is parametrized by beta and digamma functions. An asymptotically least-favorable reset family has an atom at zero below a branch junction, whereas the constructed higher-ratio family is diffuse. At every fixed interior ratio, the thermal reset is strictly less harmful than the minimax reset; the limiting loss has square-root onset and quadratic contact with its plateau. At reset-to-input energy ratio at least one the loss is the familiar binary entropy h2(lambda). The proof uses global entropy inequalities and feasible distributions rather than an unproved exchange of limit and optimization. Finite probability-game checks support the formulas but do not supply the proof. No efficient coding theorem, unknown-channel universal code or exact finite-energy capacity is asserted."
  }
]