Elamur Groupants⌕
☰

world-school-student-state-updates

The groupant beside world/src/school.rs in the Mova repository: what the code computes, what it never does, how to go back from its outputs, its invariants, cost and checks.

Facts

Implementation material. A Mova code model describes this code at one source revision: partial models, finite checks and their stated scope. It is not a reviewed general exposition, and its labels are shown as written, never raised.

Components

componentlabelwhere
approximantpartial source modelforward
astpartial../../syntax.json
compositionpartial source modelcomposition
costpartial source modelcost
descriptantpartial source modelepoch-memory
invariantpartial source modelinvariant
testspartial source modelchecks

Obernants

The typed relations around the forward model: the behaviour complement (what the code never does), the reverse reconstruction (every input compatible with an output) and specialised views.

idtypelabelwhere
world-school-student-state-updates-behavior-complementbehavior_complementpartial source modelbehavior-complement
world-school-student-state-updates-reverse-reconstructionreverse_reconstructionpartial source modelreverse-reconstruction
world-school-student-state-updates-couplingconstraint_relationpartial source modelcoupling
world-school-student-state-updates-observation-fiberinformation_losspartial source modelobservation-fiber

Model text

forward

When chosen==target, tick only. Otherwise add one to target and subtract one from chosen for each feature except singleton components of a '+' hint; add three to the full hint target even if absent from text. Tick then adds every current row to its accumulator. No division by tick count occurs.

invariant

All i64 increments and accumulated sums must remain representable: production updates are unchecked. Generic +/-1 conserves row sum; hint credit adds three. Tick creates accumulator rows even for zero vectors, influencing selection.

cost

Correction pays feature extraction, O(F log K) updates and O(K log K) tick. Correct predictions still tick all current keys. Aggregated history uses O(K) space without an exercise trace.

composition

Native 'give more' targeting Add after Sub leaves only give+more=[4,-1,0,0]; the next correct tick produces acc=[8,-2,0,0]. The real state survives codec roundtrip and answers identically. No lesson run.

epoch-memory

2026-10-30 initial school source-bound epoch. Prospective model: keep O(1) arithmetic/partition cost, widen aggregation scalars, and expose actual codecs without filesystem effects. Old native witnesses found unchecked operation, score, marginal vote, partition and duplicate-vote arithmetic; legacy serialization synthesized zero accumulators. Review found that dropping explicit zero accumulators also changed the prediction table. Both activation directions now have named tests. No speedup, approximate-error profile or general semantic proof is claimed. Unchecked update/world arithmetic retains representability premises.

checks

cargo test -p world --lib school::wave12_models -- --test-threads=1. Root evidence.md binds exact source/evidence hashes, repeated debug/release outcomes and a compiled production mutation. Named fixture domains appear in composition. Syntax is a source-bound partial projection; macros remain opaque. No semantic AST or live school execution.

behavior-complement

Subject/revision: manifest hash and byte scope. Universe: supplied finite inputs, fixed dependencies, successful allocation and the explicit arithmetic premises of this scope. Inner absent-behavior relation: On representable updates, chosen==target cannot apply correction or hint credit. Singleton hint parts cannot receive generic correction. The full hint target must receive its additional three. Overflow behavior remains unmodeled. These are source-derived constraints with finite named witnesses where indicated; external behavior remains unknown.

reverse-reconstruction

Subject/revision: manifest hash. Operator: compatible preimage/converse, holding dependencies and admissibility fixed. With known event, chosen op, hint and key existence, subtract the specified correction and prior current vectors. Without that trace many example sequences and earlier accumulators produce identical w/acc. Creation of zero keys defeats unique structural recovery. Partial relational description, not a complete inverse search or a unique causal history.

coupling

Subject/revision: manifest hash. Constraint relation on stage identities and ordered intermediate states: Native 'give more' targeting Add after Sub leaves only give+more=[4,-1,0,0]; the next correct tick produces acc=[8,-2,0,0]. The real state survives codec roundtrip and answers identically. No lesson run. Supplied native joins and source inspection are distinct evidence.

observation-fiber

Subject/revision: manifest hash. Domain: scoped inputs with the declared observation. The information-loss relation equates inputs whose observation agrees: Equal final w can arise with distinct acc and different predictions. Identical w/acc can hide distinct exercise histories, so a TSV is not training provenance. Finite tests do not establish unrestricted completeness.

This page in Markdown → https://md.groupants.elamur.ai/free/en/mova/world/src/school.rs.groupant/student-state.groupant/updates.groupant/