AxionQECC: A Degradation-Aware Control Framework for Managing Absorbing Failure in Quantum Error Correction Simulations

Abstract We present a control-oriented framework for evaluating the long-horizon survivability of quantum error-correction (QECC) systems under sustained noise and leakage. Conventional QECC evaluations emphasize short-horizon logical error rates and threshold behavior, and typically do not distinguish between recoverable degradation and irreversible failure. As a result, they provide limited insight into system-level collapse modes that arise during extended operation. The proposed framework, AxionQECC, models error correction as a supervisory control process with explicit regime states and absorbing unrecoverable failure. The approach introduces degradation metrics, boundary estimation, and regime-aware control actions designed to prevent entry into absorbing states, enabling analysis beyond instantaneous decoding performance. We implement the framework in simulation and evaluate it using extended wall-clock stress tests over a broad parameter sweep spanning noise scale, leakage scale, and leakage probability. Across 500 runs (100 regimes, five seeds each) with a fixed 400-episode horizon, we observe regime-dependent survivability behavior, including sharp transitions between stable operation and rapid absorbing collapse. This work is limited to simulation and the QECC control layer. We make no hardware claims and do not propose a new error-correcting code, decoder, or threshold improvement. Instead, the contribution is a general control abstraction and evaluation methodology for studying long-horizon failure and recovery dynamics in simulated QECC systems.