Index listing of all 4 items tagged with #models.
A trained model is a file of learned numbers that turns input into likely continuations. What it does on its own, and what the software around it supplies.
An embedding turns text into a list of numbers, so that related text lands close together. What the numbers, dimensions and similarity scores mean.
Metaphors structure how we act on ideas.
A conceptual deck on JEPA-style world modeling and why prediction targets matter for system behavior.