There's definitely something wrong with the capx.

The event sheet showed 0 events like Ubivis reported, but the preview is still logging messages to the console like you said.

At first I thought it was some kind of hard-wired behavior, but after some time adding events (in the previously empty event sheet) a lot of events popped out of nowhere.

I guess the capx is corrupted somehow.

Also the action to set the learning rate and momentum factor is wrongly named as a duplicated "Define hidden node count".

I would like to suggest an action to define/add inputs in the same way as inputs are set, so they could be dynamically defined through expressions. This would allow to run a loop to define all inputs, which would be extremely useful when using an array as input for a feature detection NN for instance. Something in the line of:

 Define new input str( loopindex("x")&loopindex("y") )[/code:1jmtvu2q]
Currently it's impractical to define all inputs manually through the "Add parameter" button, since a small 10x5 array would need 50 inputs to be defined by hand.

I was about to say the same thing as R0J0hound.

Probably what is happening is that by only considering one wall the player is pushed into the opposite wall, and on the next frame pushed further into the previous wall, and so on... so it spiral out of control.

I think one way to solve is to instead of generating a push out angle, generate a vector that represents that angle (x,y), and then you just need to sum all the vectors of all the overlapping walls to get the resultant push out angle.

I bought your plugin today and was testing it's features, but I'm having problems with getting the object matrix to work.

By seeing Tiny Tank I guess you used the object matrix to do the skew on the tank parts. So I set a shear matrix and tried both the manual and auto update, but the object doesn't show any difference. Am I missing something?

I tried a lot of different values and searched for an action or property to maybe enable the feature but couldn't find anything.