Quantum 2025 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
MON: Monday Contributed Sessions
MON 18: Quantum Algorithms
MON 18.6: Vortrag
Montag, 8. September 2025, 17:45–18:00, ZHG007
Quantum Text Generation with Quantum Context-Sensitive Word Embeddings: A Comparative Architecture and Experimental Analysis — •Charles Varmantchaonala M.1, Niclas Götting1, Nils-Erik Schütte1, J. L. E Fendji2,3, and Christopher Gies1 — 1Institut für Physik, Fakultät V, Carl von Ossietzky Universität Oldenburg, 26129 Oldenburg — 2Department of Computer Engineering University Institute of Technology University of Ngaoundéré, Ngaoundere, Cameroon — 3Stellenbosch Institute for Advanced Study (STIAS) Wallenberg Research Centre at Stellenbosch University Stellenbosch, South Africa
Quantum machine learning has recently gained attention for its potential to enhance natural language processing tasks[1,2]. In this talk, we present a quantum-based text generation architecture that incorporates a quantum-native word embedding method using parameterized quantum circuits. This approach encodes classical contextual information into quantum states by designing specific quantum circuits, resulting in word embeddings that leverage quantum properties. These embeddings are then used in a prototype text generation model. To assess its effectiveness, we perform a comparative analysis against a classical model using small-scale and controlled datasets. The talk highlights both the current limitations and the potential of quantum word embeddings in language modeling. We conclude with a discussion on outlooks toward near-term quantum language tasks.
1. C. Varmantchaonala M. et al., IEEE Access 12, 99578 (2024)
2. J. Shi et al., IEEE TNNLS, 1 (2024)
Keywords: Quantum Machine Learning; Quantum Natural Language Processing; Quantum Word Embedding; Natural Language Processing; Quantum Computing