Function py_wong_wang_simulate 

pub(crate) fn py_wong_wang_simulate<'py>(
    py: Python<'py>,
    s1_init: f64,
    s2_init: f64,
    tau_s: f64,
    gamma: f64,
    j_n: f64,
    j_cross: f64,
    i_0: f64,
    sigma: f64,
    dt: f64,
    stim1: PyReadonlyArray1<'py, f64>,
    stim2: PyReadonlyArray1<'py, f64>,
    xi: PyReadonlyArray1<'py, f64>,
) -> PyResult<Py<PyAny>>

Simulate a single Wong-Wang unit for stim1.len() steps and return the per-step state traces + firing-rate traces.

Parity contract with sc_neurocore.neurons.models.wong_wang.WongWangUnit: the caller pre-draws 2 * n_steps N(0, 1) samples so Python-side numpy.random ordering is preserved bit-for-bit.

Returns a dict with keys: s1, s2, r1, r2 (1-D float64 arrays of length n_steps) and the final scalar s1_final, s2_final.