RatLabA Boar Research
Closed recordas of what’s still open →

The map · 164 nodes · 109 links

Everything, connected

The story runs clockwise around the ring, one chapter per hub. Around each hub sit the ideas, tests and choices of that chapter; lines show what led to what, what was ruled out, and what uses which idea. The author's thesis and the on-phone headline sit in the middle. Point at or tap a node to see its neighbours; open it to read its note. On a phone, the list below is the quickest way in.

Browse the map as a list, chapter by chapter

01The question

The goal was simple to say and hard to measure, so the first job was building the ruler.

02Building a ruler

Every answer is graded twice, blind, against a frozen frontier answer, under rules written before the test.

03The floor trap

We grade on the scale where a blank answer scores zero, and still show the flattering scale beside it.

04Seven frozen builds

Seven builds made the app safer and quicker, not smarter.

05The walls

The small model reads well enough; what it lacks is knowledge the phone doesn't hold.

06The deep tier

A bigger model that thinks first got closest, 0.436, on a desktop, with no wins, and it was never shipped.

07A phone on a desk, and in a hand

The phone matches the desktop on quality; it is just slower, and the big model never left the desktop.

08The agent lab

An AI-agent lab can run real research, if every rule is written down and every mistake becomes a new rule.

09Knowledge packs

Pick the one library you need; more is not better.

10Measuring the measurer

A different judge changes the level, not the ranking, and not the wins.

11A smaller engine for 8 GB phones

For an 8 GB phone, speed and memory picked the small model; it reads worse, and we say so.

12A second ruler (bench v0.4)

A fairer exam is written and sealed; the submitted app has sat it, and the grade waits for the author (as of 24 October 2026, 20:34).

13Delivery, not coverage

The phone usually has the book; the hard part is opening it at the right page.

14Closing the research: what a small model needs

A small model needs a library: right about two times in three when the page arrives; the page arrives about half the time.

15One simple app: LabRat

The entry is a deliberately simple app, LabRat, that runs offline on a phone; as of 25 October 2026 its desktop dry run is below the bar and its final test on the phone, the one that counts, is below the bar too.

16Where it stands (live)

Not shown to be half as good: the desktop dry run of the new exam is below the bar, the phone run that counts is below the bar too, and a small model is right about two times in three when the page arrives (as of 25 October).

17Open fronts and hand-off

The research is closed; what remains is the phone test of the simple app, hosting it, and posting the claim.

18Reproduce it (appendix)