Collected sources and patterns will appear here. Add from search, explore, or the patterns library.
TimeSeries<Raw> -> TimeSeries<Normalized>
Normalize input time-series metrics by local mean and variance and restore the scale post-prediction.
Problem it solves
Observability metrics span wildly different scales (e.g., network bytes vs CPU percentages), causing gradient instability in unified foundation models.
Consumes
Emits
The real projects this mechanism was found in. Attribution is the point — this is how the best teams actually do it.