The concepts, in plain words.
The idea, the problem, and what it's meant to do. Not how it's built.
QTSCE
Quantum-Topological Siamese Cartographic Engine.
QTSCE is the name of the language network I'm building: one designed to understand the inherent language of communication — not just words, but what underlies them: cause and effect, context, direction and intent.
I don't share architecture details publicly. Technical detail is available to qualified partners under NDA.
A photograph, not a phrasebook
Most language models learn something like a huge phrasebook of what usually comes next. I'm building one that works more like a photograph of meaning.
Geometric representation means meaning captured as a continuous structure — the way brightness varies across a photograph, or loudness across a sound — rather than as statistics about word order. It's a different foundation, not a tweak to existing models.
What's wrong with the phrasebook
Today's language models are expensive to train, need huge amounts of data, are hard to inspect, and will give confident answers when they don't actually know.
Researchers, developers, and software and firmware teams need tools they can understand, check and trust.
What it's meant to do
- Aims to learn structure efficiently — including from rare examples.
- Aims to be inspectable — so you can see why it gave an answer.
- Aims to say when it doesn't have enough information — rather than guess.
Where it's at
Research prototype. Results will be published when properly tested.
I'll pre-register hypotheses and test on held-out data against baselines.
Patent pending (Canada).
Words I use, defined once
| Term | What it means here |
|---|---|
| geometric representation | Meaning captured as a continuous structure — the way brightness varies across a photograph, or loudness across a sound — rather than as statistics about word order. |
| held-out data | Data kept out of training and used only for testing. |
| baseline | A standard model to compare against. |
Who it's for
Research labs, software teams and firmware teams who need language tech they can inspect. The plan is developer API licensing, once results have been independently tested.
Want to talk about it? Email me at raffleraffle@arbloodhound.xyz.
Common questions
Can I try QTSCE?
Not yet. There's no public sandbox or API right now. Research prototype. Results will be published when properly tested.
How does it work?
I don't share architecture details publicly. Technical detail is available to qualified partners under NDA.
When will results be published?
When they've been properly tested. I'll pre-register hypotheses and test on held-out data against baselines.
What is the IP status?
Patent pending (Canada).
Who is behind this?
Just me — Rafael Cabrera, in Victoria, BC. LinkedIn · raffleraffle@arbloodhound.xyz