Re:Backpropagation is a leaky abstraction (2016)
by away74etcie on Sunday November 02, 2025 @07:28AM
(#45789820)
Karpathy's work on large datasets for deep neural flow is conceiving of the "backward pass" as the preparation for initializing the mechanics for weight ranges, either as derivatives in -10/+10 statistic deviations.
[ Parent
| on HN ]
There are no comments on this comment yet.