Unstructured parameterized quantum circuits suffer barren plateaus because their high expressivity produces exponentially flat gradients; symmetry-preserving designs that keep dynamical Lie algebra dimension polynomial act as a structural regularizer enabling scalable training.
A Geometric-Aware Perspec- tive and Beyond: Hybrid Quantum-Classical Machine Learning Methods,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Beyond the Expressivity-Trainability Paradox: A Dynamical Lie Algebra Perspective on Navigating Barren Plateaus in Quantum Machine Learning
Unstructured parameterized quantum circuits suffer barren plateaus because their high expressivity produces exponentially flat gradients; symmetry-preserving designs that keep dynamical Lie algebra dimension polynomial act as a structural regularizer enabling scalable training.