Skip to main content
T·M TOMFORMER live research instrument Patent pending

Add a document. Nothing else moves.

A document is read into TOMX's knowledge through a gate that decides, per fact, whether to write it, skip it as already known, or flag it as contradicting what is already held. No retraining: the knowledge lives on disk, so adding to it is a write.

What this shows, and what it does not

Every entity this deployment serves already holds its document, so there is no blank entity to teach. What you can do is withhold one entity's document for your session, watch its answers degrade, then paste the document back and watch them return. That is a real write through the real gate, and it is how the research run itself was measured. It is not a document about something the corpus has never seen, and this panel does not claim that.

Fixed-probe proof

Grow once. Rerun the identical old world.

First hash and run the server-owned probe500 manifest. Then add one safe, previously absent fact, answer its new query, and rerun that exact manifest. The checkpoint identity is read before and after; no-retraining passes only when it stays identical while overlay rows grow.

live driver
1 · Baseline

Clear this session overlay; run and hash the fixed probe.

2 · Grow + ask

Write the fixed safe document and answer the new fact.

3 · Verify

Rerun the same probe and show every old-world delta.

probe500 · 500 rows · TOMX · overlay empty at start

Germany

Q183
document held
Relation What the model answers now Margin Graph says
P17- Cologne 0.1072 Alsfeld
P27- Western Europe 0.2935 Horst Bienek
P530- Russia 0.1134 Australia
P495- Austria 0.0214 Southland Tales
P150 Austria 0.0214 Hamburg
P915- history of Europe 0.0059 Chitty Chitty Bang Bang
P463 Austria 0.0214 Eurocontrol
P530 Russia 0.1134 Taiwan

Paste the document

Facts are recognized only between entities this corpus already holds, joined by a relation it already knows. This is not an open-domain extractor. When it finds nothing it reports that instead of inventing a fact.

up to 8000 characters

Does adding break what was there?

Each point is a real bench run over the probe set under your session's state: relational-hop accuracy, gate off. If growth damaged what was already held, these bars would fall. Each number is measured fresh when you press the button.

no measurements yet; take a baseline before you grow, then again after

CHANGE · GROW

Shows: a new document read into its knowledge, with the gate's decision per fact including contradictions against what is already held. Does not show: extraction quality. The mock recognizes only the sentence forms this corpus is stated in.