Quantum error correction (QEC) is often implemented on hardware with biased noise, where Pauli-$Z$ errors occur more frequently than Pauli-$X$ errors. This has motivated many recent efforts to develop bias-tailored QEC codes, such as the XZZX surface code. In this work, we investigate the Clifford-deformed compass codes, a family of codes that achieve high thresholds at large noise bias. We perform circuit-level simulations of the Clifford-deformed elongated compass codes under a hardware realistic biased noise model and evaluate code thresholds using standard Minimum Weight Perfect Matching (MWPM) and Pymatching's correlated MWPM. We find that correlated decoding significantly enhances thresholds by up to 35% relative to standard MWPM. Our results demonstrate that correlated decoding can improve the performance of Clifford-deformed asymmetric QEC codes in realistic noise regimes.
