NEST 3.9.0
NESTML 8.2.0
Python 3.12.3

I am using the iaf_psc_delta example neuron and find that membrane dynamics and spike timings deviate between NEST and NESTML.

Setup: one Poisson neuron -> two parrot neurons -> each parrot is connected either to one NEST iaf_psc_delta neuron or one NESTML iaf_psc_delta_neuron.nestml neuron.

See debug.py

Expectation: both neurons should have exactly same membrane dynamics and spike timings.
Finding: the NESTML model ignores the absolute refractory period -- V_m still integrates spikes even though this code in iaf_psc_delta_neuron.nestml should prevent that:

    update:
        if refr_t > 0 ms:
            # neuron is absolute refractory, do not evolve V_m
            integrate_odes(refr_t)
        else:
            # neuron not refractory
            integrate_odes(V_m)

The same does not happen if I use a constant current I_stim.