<ul><li>Researchers have developed a hybrid classical-quantum doubly stochastic Transformer (QDSFormer) which incorporates a variational quantum circuit in place of the Softmax in the self-attention layer.</li><li>The QDSFormer yields more diverse doubly stochastic matrices (DSMs) that better preserve information compared to classical operators.</li><li>In multiple small-scale object recognition tasks, the QDSFormer outperforms a standard Vision Transformer and other doubly stochastic Transformers.</li><li>The QDSFormer shows improved training stability and lower performance variation, potentially mitigating the unstable training of Vision Transformers on small-scale data.</li></ul>

Quantum Doubly Stochastic Transformers

Discover more