kalomaze: Modern DNI May Approximate the Regimes Where Backprop Fails

kalomaze · x · 2026-09-30

kalomaze argues that ES/genetic-algorithm training is "Sakana-tier nerdsnipe" — if you abandon backprop, do it for regimes where naive backprop is intractable without BPTT. His take: there's an unobservable regime we can't train directly; DNI loses to full backprop where the latter is tractable but behaves close to it; and a modern, especially distributional, form of DNI could yield a fairly faithful approximation of that unobservable regime.

Related event: Researcher questions evolutionary training routes, backs DNI for cases beyond backprop(2 posts)→

Original post →

More from Research

Research channel →