You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This came up in Terry's SPA tutorial on Friday. One of the most helpful things when debugging a SPA network is seeing which actions are being selected by the Basal Ganglia and Thalamus. At this point when plotting the values, you get the standard plot with no labels. It's hard to translate those colours into selected values. A legend or maybe an entirely new plot would be helpful.
The text was updated successfully, but these errors were encountered:
Also, just like you can click on the SPA related inputs to get a semantic pointer without expanding the network, I would like to be able to get the output of the basal ganglia and thalamus without expanding the network.
This would definitely still be handy. In recent debugging of SPA modules, I've also found it handy to have these be cloud plots (so that it's easy to see what actions are being selected -- if there's more than 6 actions the coloured lines aren't all that useful).
This came up in Terry's SPA tutorial on Friday. One of the most helpful things when debugging a SPA network is seeing which actions are being selected by the Basal Ganglia and Thalamus. At this point when plotting the values, you get the standard plot with no labels. It's hard to translate those colours into selected values. A legend or maybe an entirely new plot would be helpful.
The text was updated successfully, but these errors were encountered: