I was wondering if lucid can be applied to sequential models that use tabular data to visualise activation maps throughout the network based on how different layers/nodes in layers interact with one another and where they lie in terms of dimensionality reduction? if this is a possible is there an example of how to apply this?
Hi all,
I was wondering if lucid can be applied to sequential models that use tabular data to visualise activation maps throughout the network based on how different layers/nodes in layers interact with one another and where they lie in terms of dimensionality reduction? if this is a possible is there an example of how to apply this?
many thanks!