Lab notes
Mind-1B is a brain built from a billion ternary parts. These are its lab notes: one entry when a milestone gate passes or fails, grounded in a prototype you can run. No schedule, no filler, and the open bets stay named. Honesty over hype is the house rule.
A model that never sleeps forgets its first task by the time it learns its third. We gave the prototype a wake and sleep cycle: replay the day, downscale the noise, prune what did not matter. Recall held at 81% across eight cycles, against 34% without it.
Read the note →A transformer spends the same compute on a comma as on a proof step. We wired one prediction-error signal to the throttle: on a redundant stream the prototype recomputed 7% of the dense FLOPs per token and produced the same output. Compute should follow surprise.
Read the note →The transformer is a magnificent, uniform machine, and the uniformity is the tax. It runs every region at once, grows its memory without bound, and multiplies floats it could skip. A brain does none of those things. Here is the case for a different shape.
Read the note →One email when a gate passes or fails. Written by the lab, read by humans.