arXiv:2611.13985v1 Announce Type: new Abstract: With the rise of neuromorphic technologies, simple and tunable neuron models are needed to analyse and design spiking systems. However, the simplest neuron models, e.g., Leaky Integrate-and-Fire (LIF) models, usually fail to capture the excitable behaviours of biological neurons. This could be a significant limitation, as excitability enables reliable spike timing under parametric uncertainties. Recently, a minimal excitable model, the so-called High-Pass Model (HPM), was introduced in neuroengineering. It simply consists of a static sigmoid with a high-pass filter in positive feedback. Given its novelty, it is still unclear how easily the HPM can be tuned, and how much of the excitable behaviours of a biophysical neuron it can reproduce. To address these questions, we investigate the tunability of a first-order HPM sub- and supra-rheobase, i.e., below and above the onset of the tonic spiking regime (where the neuron spikes periodically), and compare its behaviours with those of a simple Hodgkin-Huxley-type Conductance-Based Model (CBM). Below rheobase, both models fire spikes in response to transient inputs, including post-inhibitory rebound spikes. Above rheobase, an affine reparametrization of the input of the HPM suffices to fit its frequency-current curve to that of a CBM. The two models mostly differ near their respective rheobases, where the first-order HPM lacks bistability and spike latency. We suggest that modifying the frequency-selective element of the HPM may be sufficient to overcome this limitation. These results highlight the potential of the HPM as a fundamental computational primitive in neuromorphic, spike-based technologies.