Stochastic excursions: from counting observables to inference

Título: Stochastic excursions: from counting observables to inference Resumo: We introduce a framework to analyze the statistics of counting observables accumulated over stochastic excursions, which are trajectory segments with clear physical meaning. The framework provides analytical expressions for currents, heat, entropy production, and activity, capturing their correlations and finite-time fluctuations. We show how distinct noise sources contribute to these fluctuations, and prove a fluctuation theorem for excursions that implies a precision-dissipation trade-off. We then show how excursions enable thermodynamic inference, and illustrate the framework on two systems: cooling in the three-qubit absorption refrigerator, and a double quantum dot, where we infer properties of the underlying dynamics directly from experimental trajectories.