WiMi builds quantum network for better data sorting

WiMi Hologram Cloud Inc. (NASDAQ: WiMi) is applying a quantum convolutional neural network, built with three-qubit interactions, to classify classical data. The company’s new network aims to improve expressive power and entanglement generation, key to advancing quantum deep learning model design. WiMi’s approach utilizes block partitioning for image data and combines structured amplitude and angle encoding for one-dimensional data, efficiently translating classical information into a quantum format. Three-Qubit Interactions Enhance Quantum Convolutional Neural Network Expressivity WiMi Hologram Cloud Inc. This approach deviates from the typical focus of quantum computing on problems intractable for conventional computers, instead targeting established machine learning tasks with a novel quantum architecture. A core innovation lies within the network’s interaction layers, specifically designed around three-qubit interactions. WiMi researchers systematically studied how these layers impact the quantum state space coverage, finding that introducing three-body interactions significantly expands the range of reachable states within the network’s parameter space. This expansion directly addresses a common limitation in traditional quantum neural networks, a lack of expressive capacity for complex patterns, while simultaneously maintaining manageable circuit depth.

The team further analyzed the network’s entanglement capabilities through quantum information theory, revealing that the three-qubit interaction layer generates high-intensity, multi-scale entanglement at relatively shallow depths, which is crucial for capturing nonlinear correlations within input data and offers a clear advantage over models relying solely on two-qubit entanglement gates. For image data, WiMi employs block partitioning and local mapping to embed pixel information into quantum subsystems, while one-dimensional data benefits from a combination of structured amplitude encoding and angle encoding, the company says. This efficient data encoding strategy preserves crucial discriminative information from the original data, even with limited qubit resources. After encoding, the quantum state enters a feature extraction module comprised of alternating quantum convolutional layers and the newly designed interaction layers. Quantum convolutional layers extract low-order features within local qubit subspaces, adhering to hardware-friendly principles to minimize complex gate sequences. WiMi states that “by deeply integrating multi-qubit interaction mechanisms at the network structure level, this technology provides solid support for performance breakthroughs of quantum convolutional neural networks in practical applications,” highlighting the potential for real-world impact. The network’s training utilizes a joint iterative process between classical optimizers and quantum circuit parameters, with measurement mapping quantum circuit outputs to classical feature vectors evaluated by a classical loss function. WiMi optimized parameter initialization and training procedures to address gradient vanishing and instability, achieving stable convergence in both multi-class and binary classification tasks, according to the company. This focus on practical deployability, alongside theoretical advancements, signals a shift toward quantum-native intelligent models, potentially unlocking disruptive computational capabilities as quantum hardware continues to evolve. Source: https://www.prnewswire.com/news-releases/wimis-next-generation-quantum-convolutional-neural-network-reshapes-classical-data-classification-methods-302857369.html Stay currentSee today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: The Neuron With a keen intuition for emerging technologies, The Neuron brings over 5 years of deep expertise to the AI conversation. Coming from roots in software engineering, they've witnessed firsthand the transformation from traditional computing paradigms to today's ML-powered landscape. Their hands-on experience implementing neural networks and deep learning systems for Fortune 500 companies has provided unique insights that few tech writers possess. From developing recommendation engines that drive billions in revenue to optimizing computer vision systems for manufacturing giants, The Neuron doesn't just write about machine learning—they've shaped its real-world applications across industries. Having built real systems that are used across the globe by millions of users, that deep technological bases helps me write about the technologies of the future and current. Whether that is AI or Quantum Computing. Latest Posts by The Neuron: NSF plans $280M for biology research, down from $640M in 2023 August 21, 2026 Hartree Centre and FormationQ will help UK firms use quantum tools August 21, 2026 SandboxAQ’s new tool ranks drug candidates for just $1 per 1,000 August 20, 2026