Concept · 2 episode(s)

Iterative Training

← all concepts

Definition

Iterative training alternates between training a model and generating new training data — from the model itself, from its environment, or from improved labels — rather than training once on a fixed dataset. It’s how a lot of modern reasoning models are made.

Episodes covering this

Worth reading next

Papers we haven't done a deep dive on yet, but would recommend on this topic.