Angle-encode a two-feature datum

The circuit declares two non-trainable data symbols $x_0$ and $x_1$ (parameters named x0 and x1). Build the angle-encoding feature map

$\lvert\phi(x)\rangle=R_Y(x_0)\lvert0\rangle\otimes R_Y(x_1)\lvert0\rangle=\Big(\cos\tfrac{x_0}2\lvert0\rangle+\sin\tfrac{x_0}2\lvert1\rangle\Big)\otimes\Big(\cos\tfrac{x_1}2\lvert0\rangle+\sin\tfrac{x_1}2\lvert1\rangle\Big)$

with one $R_Y$ per qubit: qubit $0$ is rotated by the symbol x0 and qubit $1$ by the symbol x1. Use the symbols as the gate angles, not numbers: the circuit must be a function of the datum.

How it is graded. The grader substitutes the three data points $(x_0,x_1)\in\{(0.3,1.2),\,(1.0,0.4),\,(2.5,-0.7)\}$ for the symbols, runs your circuit for each, and compares, with absolute tolerance $10^{-6}$:

• check z: $\langle Z_0\rangle=\cos x_0$ and $\langle Z_1\rangle=\cos x_1$;
• check x: $\langle X_0\rangle=\sin x_0$ and $\langle X_1\rangle=\sin x_1$.

For example, at the first datum the targets are $\langle Z_0\rangle=\cos0.3=0.955336$ and $\langle X_0\rangle=\sin0.3=0.295520$. Only $R_X$, $R_Y$, $R_Z$ are available, at most two gates and no two-qubit gates; the parameter list must stay x0, x1 with their values unchanged.