QuICS Special Seminar: Pengyu Liu

Title:  Achieving Optimal-Distance Atom-Loss Correction via Pauli Envelope
Speaker:  Pengyu Liu (Carnegie Mellon University)
Date & Time:  April 8, 2026, 3:30pm
Where to Attend:  ATL 3100A and Virtual Via Zoom: https://umd.zoom.us/j/7031888774?omn=91328805747

Atom loss is a major error source in neutral-atom quantum computers, accounting for over 40% of the total physical errors in recent experiments. Unlike Pauli errors, atom loss poses significant challenges for both syndrome extraction and decoding due to its nonlinearity and correlated nature. Current syndrome extraction circuits either require additional physical overhead or sacrifice loss tolerance. Existing loss decoders are computationally inefficient, achieve suboptimal logical error rates, or rely on machine learning without provable guarantees. To address these challenges, we propose the Pauli Envelope framework, which bounds the effect of atom loss with low-weight, efficiently computable Pauli approximations, generalizing existing loss-to-Pauli methods and enabling rigorous analysis. Guided by this framework, we first design a new atom-replenishing syndrome extraction circuit, the Mid-SWAP syndrome extraction, that reduces error propagation on rotated surface codes at no additional space-time cost. We then propose an Envelope-MLE decoder for Mid-SWAP syndrome extraction, formulated as a Mixed-Integer Linear Program (MILP), achieving the optimal loss distance (d_{loss} \sim d). Inspired by its exclusivity constraint, we also propose an Envelope-Matching decoder that approximately enforces the constraint within Minimum-Weight Perfect Matching (MWPM), achieving (d_{loss} \sim 2d/3) and surpassing the previous best algorithmic decoder ((d_{loss} \sim d/2)). When combined with recent fast correlated-decoding techniques, our Envelope-Matching decoder enables efficient, high-threshold loss decoding for transversal surface-code logical circuits. Circuit-level simulations demonstrate up to 40% higher thresholds and 30% higher effective distances compared with existing methods in the loss-dominated regime. Moreover, we explore the problem of correlated atom loss and show that it is easier to correct than independent loss, with the loss threshold rising from 5.15% to 7.82% as the correlated fraction increases. Remarkably, our Envelope-MLE decoder improves the error suppression factor of a hybrid MLE–machine-learning decoder from (\Lambda = 2.14) to (\Lambda = 2.24) on recent experimental data.

Bio:  Pengyu Liu is a third-year PhD student in Computer Science at Carnegie Mellon University, advised by Prof. Umut Acar. His research interests lie in quantum error correction and quantum computer architecture.