Are current decay and voltage decay learnable using spike time loss?

[JDRanpariya]

I have a small experiment where I'm trying to learn the spiking behavior "making a single neuron learn to spike if it sees two consecutive spikes in input", It works fine when synaptic weight is learnable but when I freeze it and make current and voltage decay learnable, they change very gradually not much and neuron doesn't learn the expected behavior.

does spike time loss change them or they are not trainable? I'm currently using learning rate of 0.01

Best,
Jay