aiqbench analysis
This breakthrough demonstrates a high-fidelity leakage reduction unit that actively corrects qubit errors during measurement without introducing time overhead, directly addressing a major bottleneck in scaling superconducting quantum computers. By integrating this hardware design with neural-network decoding, the research significantly suppresses logical error rates while maintaining readout fidelity, marking a critical step toward practical fault-tolerant quantum computation.
Quantum error correction avoids pausing with new hardware design

Leakage to non-computational states is a source of correlated errors limiting quantum error correction (QEC) with superconducting circuits, and current fixes require pausing the quantum computer, wasting time. Researchers present and experimentally demonstrate a high-fidelity leakage reduction unit (LRU) operating concurrently with transmon measurement without incurring time overhead for 2-level readout.
Yuejie Xin and colleagues have demonstrated a leakage reduction unit (LRU) achieving 98.4% leakage removal during qubit measurement without adding time overhead, a critical step for improving quantum error correction. Adapted from a protocol called DDROP, the LRU utilizes simultaneous drives on the transmon and its readout resonator, actively returning qubits to a usable state. The LRU achieves a 98.4% leakage removal fraction without compromising the 2RO assignment fidelity (99.2%). The team combined this enhanced measurement with neural-network decoding to successfully suppress logical error rates in experiments with up to nine qubits. A 98.4% leakage removal fraction was experimentally achieved without adding any time overhead to the standard two-level readout process, a critical advancement for scaling quantum error correction in superconducting circuits. This improvement stems from a newly developed Leakage Reduction Unit (LRU) integrated directly into the qubit measurement process, circumventing the need for additional processing steps that typically slow down computations. The LRU operates by simultaneously applying microwave pulses to both…
Excerpt shown under the publisher's RSS terms · 162 words · full text and figures are on the publisher's site
Related coverage

NVIDIA’s NVLink 6 keeps AI factories running through any error
Quantum Zeitgeist · Sep 16, 2026

The Two-Century Path That Made Quantum Computing Inevitable
Quantum Computing Report · Sep 28, 2026
Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
Latent Space · Oct 8, 2026

Gemini 4 Argon: Built to solve your hardest problems
Google DeepMind · Oct 13, 2026