Fig. 9: Average results under restricted Haar ensemble.
From: Quantum-data-driven dynamical transition in quantum learning
We plot (a) Kαα(∞) versus y1 with y2 = 0.5 and L = 256 fixed, (b) λααα(∞) versus L with y1 = 5, y2 = 6 fixed. Blue and red dashed lines in (a) represent Eq. (43). Blue and red dashed lines (overlapped) in (b) represent Eq. (44). Here random Pauli ansatz (RPA) consists of L variational parameters on n = 4 qubits. There are N = 2 orthogonal data states and the corresponding target states are computational basis \(\left\vert 0000\right\rangle ,\left\vert 0001\right\rangle\).
