Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Final output values do not make mathematical sense. #194

Open
jeffreywolberg opened this issue Aug 18, 2023 · 2 comments
Open

Final output values do not make mathematical sense. #194

jeffreywolberg opened this issue Aug 18, 2023 · 2 comments

Comments

@jeffreywolberg
Copy link

I have the following playground that I was tinkering with:

https://playground.tensorflow.org/#activation=relu&batchSize=10&dataset=circle&regDataset=reg-plane&learningRate=0.03&regularizationRate=0&noise=15&networkShape=3&seed=0.97439&showTestData=false&discretize=false&percTrainData=50&x=true&y=true&xTimesY=false&xSquared=false&ySquared=false&cosX=false&sinX=false&cosY=false&sinY=false&collectStats=false&problem=classification&initZero=false&hideText=false

After 100 epochs, I get the following decision boundary

image

However, I notice something weird. Notice the output of the hidden layers in the above pic. They are all white/blue (>= 0, which makes sense because I am using ReLU). However, all of the 3 final weights are negative. This must mean that all of the output is negative (since positive inputs multiplied with negative weights must be negative). However, in the outputted decision boundary shown, there's plenty of values >=0 (in blue). How does this make sense? Is there normalization or some bias being added to the output neuron? If so, why is it now shown in the diagram?

@AngelODeath
Copy link

What browser version and OS version are you using?

@piojanu
Copy link

piojanu commented Sep 22, 2023

@jeffreywolberg see #90. The output node in deed has a bias as any other node. Moreover, the output activation for the classification task is tanh. This explains a lot. This should be included in the playground description somewhere.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants