Researchers study an adversarial bandit problem for entanglement-based quantum-network routing over a modest graph corpus, where Alice selects routes for the Ekert-91 protocol and Eve selects attack surfaces.

  • Alice accepts a turn when the finite-sample statistic violates the Clauser-Horne-Shimony-Holt (CHSH) bound.
  • Learned retention tracks a full-matrix minimax reference closely with a Pearson correlation of 0.99 across 50 structured topologies.
  • Under a one-surface Eve action model, bottleneck families have zero retention while non-bottleneck families follow a 1-1/N coverage principle.
  • Decision-tree explanation models are fitted to graph-, attack-, and route-level targets to report faithfulness.
  • Prompt records for local language models summarize tree evidence, creating an open-source explanation workflow for quantum-repeater network games.

The work provides an open-source explanation workflow for quantum-repeater network games by summarizing decision-tree evidence using local language models.