A small model that
loads one specialist
per question.
Mnema V1 has 360M parameters. Each skill is a sealed specialist, and a router picks the one that fits your question. You can ask it something now.
What is a masked-diffusion scheduler?
The skill is
in a sealed unit.
A specialist ships as one .ncn file. Nothing about it is guessed at load time: the file says what it is for and what it is not for.
Adding a domain means adding a file. The base is not retrained.
- Weight delta rank 16
The part that changes the base's behaviour.
- Capability capsule what the router reads
The vector each question is scored against.
- Scope and exclusions declared
What the specialist is for, and what it is not for.
- Validation set bound to the file
The checks it has to pass travel with it.
- Provenance record who made it
Where the specialist came from.
- 01Embed once
The question is embedded a single time.
- 02Score every capsule
Cosine similarity against each specialist's capsule, in closed form. The scores are visible per query.
- 03Attach and answer
The top scorer is attached to the base, and the base answers with it.
Selection is clean at two,
and breaks at three.
With two specialists the router cleared its 85% gate: 18 of 20 queries across two bases, both misses low-margin. With three or more specialists from similar domains, top-1 accuracy fell to 23.3% and stayed under 47% out to six. We tried centring the capsules, enlarging one corpus and replacing closed form with a learned router. Each made it worse or changed nothing.
We published that as a failure and built the next model around it. XE5 does not select. It blends every specialist on every token and takes its weights from a memory descent. The study that measured the break is XE4-S, and the model it led to is XE5.
Ask it something,
and watch the route.
The live demo shows the six scores first, then the specialist loading, then the answer as it is made.