Post #2450311
2026-04-25 01:53 UTC
@rk@mastodon.well.com The distribution of float values a computer might process in a day. Not sure why it would be Zipf- or Benford-like.
Replies (1)
-
@BoydStephenSmithJr@hachyderm.io 2026-04-25 02:22
@steven_pigeon@mstdn.ca @rk@mastodon.well.com There's a criteria https://en.wikipedia.org/wiki/Benford's_law#Benford's_law_compliance_theorem though I don't know why it would apply to "inputs to a FPU" in general -- it seems like you'd have to assume (or prove) something about what is generating those inputs (the workload or programs).