ml-pokedex About · The Arc

The Arc is where the collection started: the RDF tinyshakespeare. Train a 2.86M model from scratch and, instead of scoring a final number, watch RDF structure appear in layers, each rung lighting roughly in order.

The six rungs

Every epoch, from the same fixed prompt and seed, the model is scored on six things that should arrive in sequence:

rungwhat it meanslights
L1 syntaxemits parseable Turtle at all~epoch 5
L2 vocabularyIRIs are real ontology terms, not gibberish~epoch 5
L3 schemahasType takes a PokéType; ranges respected~epoch 5
L4 shapea Species gets the canonical field setearly, climbing
L5 reasoningweakTo is the chart's verdict on its own types~epoch 15
L6 realityknows pikachu is real and zarquon isn't~epoch 5

Structure first, meaning last

Grammar, vocabulary and schema finish within the first few epochs; the sample stops being garbage bytes and becomes well-formed triples fast. Reasoning is the slow one. The weakTo lines stay random long after everything around them is correct, and only lock onto the type chart around epoch 15 (0.000 to 1.000). The words are right long before the meaning is.

One detail worth noting: L6 (reality) lights early, with the vocabulary rather than the reasoning. Knowing which names the world contains is a property of names, and names are what byte-BPE learns first.

What you're scrubbing through

These are real samples captured during training, one per epoch, same prompt, same seed (temperature 0.7), so the only thing changing as you drag the slider is the model itself. The final epoch is not a picture: it is the same weights running live in the Minter and Reasoner. The Arc shows how they got that way.

tinyshakespeare doesn't reason either. It babbles plausible verse, and that is the whole claim it makes. This is the RDF version: a small model doing well inside a small, closed world, its abilities layering in an order you can watch.

← Back to the collection about