In a pre-print research paper titled “QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer” and posted on arXiv, Cambridge Quantum Computing (QLP) they were able to show how they can create representations for sentences that have a natural mapping to quantum circuits. The paper describes two datasets consisting of 130 and 105 sentences that were converted to representations in quantum circuits. The experiments were performed on a 5 qubit IBM NISQ quantum computer, so the algorithms could also be simulated on a classical computer. But as the complexity and scope of the data sets get scaled up, the approach may one day be used to provide a means for performing a Natural Language Processing (NLP) task on a quantum computer that would not be possible on a classical computer. For more about this experiment, you can read a news announcement available on the CQC website here and also the arXiv pre-print paper here.
March 3, 2021