AI & formal methods
What happens when Jev meets Lean?
A probabilistic model can propose a decision. A formal evaluator can make the acceptance rule explicit and reproducible.
Matha Labs · Research notes
From a model’s interpretation to an explicit decision: five open experiments in evidence, formal rules and reproducible results.
Research prototypes · 15 November 2026
AI & formal methods
A probabilistic model can propose a decision. A formal evaluator can make the acceptance rule explicit and reproducible.
Healthcare
A small ICD-10-CM research checker connects proposed codes to explicit evidence and keeps uncertainty visible.
Healthcare economics
Versioned payer assumptions, integer arithmetic and explicit prerequisites make a small payment model inspectable.
Cities & formal methods
Start with jurisdiction. A Lean research model separates Los Angeles city intake from unincorporated county intake.
Federal tax & formal methods
A Lean model of the 2026 ordinary income-tax rate schedule makes filing status, source revision and bracket arithmetic explicit.
A model or reviewer proposes structured facts and points to supporting evidence.
Lean checks the encoded conditions against fixed evidence and a named policy revision.
The evaluator returns the same result for the same inputs, including when review is required.
These releases contain executable synthetic examples and proofs about their encoded rules. Clinical deployment, payer adjudication, municipal approval and tax return preparation require additional data, validation and authorized review.