A new QNN architecture with unified graph, HAL, and ONNX pipeline enables cross-framework and cross-hardware QML with training time within 8% of native implementations and identical accuracy on Iris, Wine, and MNIST-4 tasks.
Variational quantum circuits for deep reinforcement learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.ET 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Eliminating Vendor Lock-In in Quantum Machine Learning via Framework-Agnostic Neural Networks
A new QNN architecture with unified graph, HAL, and ONNX pipeline enables cross-framework and cross-hardware QML with training time within 8% of native implementations and identical accuracy on Iris, Wine, and MNIST-4 tasks.