Realizing the full potential of quantum computing requires large-scale quantum computers capable of running quantum error correction (QEC) to mitigate hardware errors and maintain quantum data coherence. While quantum computers operate within a two-level computational subspace, many processor modalities are inherently multi-level systems. This leads to occasional leakage into energy levels outside the computational subspace, complicating error detection and undermining QEC protocols. The problem is particularly severe in engineered qubit devices like superconducting transmons, a leading technology for fault-tolerant quantum computing. Addressing this challenge requires effective multi-level quantum system readout to identify and mitigate leakage errors. We propose a scalable, high-fidelity three-level readout that reduces FPGA resource usage by $60 \times$ compared to the baseline while reducing readout time by $20 \%$, enabling faster leakage detection. By employing matched filters to detect relaxation and excitation error patterns and integrating a modular lightweight neural network to correct crosstalk errors, the protocol significantly reduces hardware complexity, achieving a $100 \times$ reduction in neural network size. Our design supports efficient, real-time implementation on off-the-shelf FPGAs, delivering a $6.6 \%$ relative improvement in readout accuracy over the baseline. This innovation enables faster leakage mitigation, enhances QEC reliability, and accelerates the path toward faulttolerant quantum computing.

Quantum Error Correction (QEC) protects qubits against bit-and phase-flip errors in the |0⟩ /|1⟩ subspace, but physical qubits can also leak into higher energy levels (e.g., |2⟩).Leakage is especially harmful, as it corrupts all subsequent syndrome measurements and can spread to neighboring qubits.Detecting leakage on data qubits is particularly challenging, since they are never measured directly during QEC cycles.Prior work, such as eraser [43], addresses this by inferring leakage from syndrome patterns using a fixed heuristic.However, this approach often misclassifies benign syndromes, triggering excessive leakage-reduction circuits (LRCs).Because LRCs are themselves noisy and slow, these false triggers lengthen QEC cycles and inflate logical error rates.We propose gladiator, a general and adaptable leakage speculation framework that works across surface code, color code, and qLDPC codes.Offline, gladiator builds a code-aware errorpropagation graph calibrated to device data.Online, it classifies each syndrome in a few nanoseconds and schedules LRC only when the observed pattern is provably leakage-dominated.This precise speculation eliminates up to 3× (and on average 2×) unnecessary LRCs, shortens QEC cycles, and suppresses false positives at their source.Evaluated on standard fault-tolerant benchmarks, gladiator delivers 1.7×-3.9×speedups and 16% reduction in logical error rate, advancing the efficiency of fault-tolerant quantum computing.

Reading a qubit is a fundamental operation in quantum computing. It translates quantum information into classical information enabling subsequent classification to assign the qubit states '0' or '1'. Unfortunately, qubit readout is one of the most error-prone and slowest operations on a superconducting quantum processor. On state-of-the-art superconducting quantum processors, readout errors can range from 1--10%. These errors occur for various reasons - crosstalk, spontaneous state transitions, and excitation caused by the readout pulse. The error-prone nature of readout has resulted in significant research to design better discriminators to achieve higher qubit-readout accuracies. High readout accuracy is essential for enabling high fidelity for near-term noisy quantum computers and error-corrected quantum computers of the future.