Decoder weights when using probe

I am learning Nengo and was going through this simple tutorial.

It is basically a model with a single neuron trying to replicate the sine wave input.

My understanding is that the input is converted into the input current with the gain and the bias current of the ensemble, which then is used to generate the spikes. Then we connect the probe with the postsynaptic filter (or low pass filter) to measure the final output of the ensemble.

My question is that how Nengo determines the amplitude of the spikes before the synaptic filter? (i.e. The voltage of the spikes from the somma of the nueron before it reaches the neuro-transmitter). I guess there must be some kind of weights for this but I am not sure how to access them.

I did more play-around with Nengo and eventually got what I wanted.
What I had to do was adding a node that connects to the ensemble then I could access the decoding weights through the connection.

But since the “probe” is essentially a node with a synapse filter, i thought Nengo must calculates the values in the same way as it would when it is connected to a node. So I thought there must be the weights calculated and stored somewhere even when you are looking at the output through the “Probe”. All the tutorials and the documents I read to access the weights are through the connection.

Is there way to access the decoding weights when you are reading the output through the probe on the ensemble directly?

Unfortunately, no. The “decoders” that are made for the Nengo probes are stored internally, and do not have any external handles that can be used to get at them. It is theoretically possible to hack the Nengo code (if you know the exact signals you are looking for), but it’s not a robust solution, and prone to errors.

The preferred method is to (as you mentioned) create a nengo.Node and access the decoders through the weights of the connection. If you set the connection between the ensemble and node to None, you can also just probe the node instead of probing the ensemble.