Researchers have developed a new method to optimize stabilizer state preparation in fault-tolerant quantum computing. This technique, based on integer linear programming (ILP) and the circuit gauge operator formalism, enables the construction of circuits with an equal or lower gate count than current methods, while detecting up to three errors. Efficient preparation of these states is crucial for logical qubit initialization and for logical-ancilla-based error correction gadgets, such as Steane and Knill codes, leading to reduced solution time and increased reliability of quantum computers.

Traditionally, for small, low-distance codes like the [[7, 1, 3]] Steane code, circuits with low gate counts could be designed by inspection. However, this approach becomes impractical for larger codes, necessitating automation. Current state-of-the-art methods for automated fault-tolerant state preparation include SAT-based stabilizer measurement and the flag-at-origin technique. The new approach enhances these methodologies by reformulating the construction of flag circuits as an integer linear programming problem, leveraging the circuit gauge operator formalism.

As a proof of concept, the technique was applied to derive a Steane error correction gadget for the [[24, 10, 4]] two-block group algebra code. This gadget was tested on Quantinuum's System Model H2 quantum computer, using 10,000 shots. The results showed a logical block error rate of approximately 0.00014 (or 0.000014 per logical qubit), with about 1.6% of the shots post-selected due to weight-two errors. These findings demonstrate the effectiveness of the proposed method in improving the robustness of quantum operations.