Collected sources and patterns will appear here. Add from search or the patterns library.
(Tensor<TimeValues>, Tensor<AttentionMask>) -> Tensor<NormalizedTimes>
Scale raw absolute timestamps using sequence limits and attention masks to produce bounded temporal intervals.
Problem it solves
Highly variable raw clinical timestamps cause numerical instability in continuous-time modeling.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.