WiMi said it has deployed a quantum-control optimization method that uses multi-objective deep reinforcement learning to improve gate fidelity, operating efficiency, noise suppression and energy use in noisy quantum systems.
The approach reuses single-process optimization results as thresholds and reward-function inputs, cutting redundant computation and speeding convergence toward a global control solution rather than a single-metric local optimum.
By modeling qubit dynamics in real time, the system automatically adjusts external-field control strategies to counter environmental noise, crosstalk and decoherence during state preparation, gate operations and readout.
The update extends a string of WiMi quantum and AI research announcements since February; shares rose 6.62% on the day, though the stock remained below its 200-day moving average of 2.87 and 52-week high of 5.65.
WiMi touts an AI solution for quantum control, but where is the proof it outperforms existing methods?
As AI begins to master quantum systems, what are the real-world implications of this powerful combination?
Beyond the hype, can an AR company truly challenge quantum giants like D-Wave and IonQ?