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
{{ message }}
This repository has been archived by the owner on Sep 1, 2023. It is now read-only.
Shouldn't that be totalOverlap instead of overlapsPredictedActive?
Otherwise it will only activate with predictedActive input and apply the weighting only in case of learning.
I know this is not a supported repo, but I fixed it in my implementation so maybe it helps someone.
The text was updated successfully, but these errors were encountered:
# for freeto subscribe to this conversation on GitHub.
Already have an account?
#.
htmresearch/htmresearch/algorithms/union_temporal_pooler.py
Line 212 in 6a823ff
Shouldn't that be
totalOverlap
instead ofoverlapsPredictedActive
?Otherwise it will only activate with predictedActive input and apply the weighting only in case of learning.
I know this is not a supported repo, but I fixed it in my implementation so maybe it helps someone.
The text was updated successfully, but these errors were encountered: