world-tune-selection
The groupant beside world/src/tune.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
| component | label | where |
|---|---|---|
| approximant | partial source model | forward |
| ast | partial | ../syntax.json |
| composition | partial source model | composition |
| cost | partial source model | cost |
| descriptant | partial source model | epoch-memory |
| invariant | partial source model | invariant |
| tests | bounded native and inspected | ../model.md#evidence |
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.
| id | type | label | where |
|---|---|---|---|
world-tune-selection-behavior-complement | behavior_complement | partial source model | behavior-complement |
world-tune-selection-reverse-reconstruction | reverse_reconstruction | partial source model | reverse-reconstruction |
world-tune-selection-tie-stability | ordering_relation | partial source model | tie-stability |
world-tune-selection-threshold-partition | acceptance_partition | partial source model | threshold-partition |
Model text
Subject: world/src/tune.rs SHA-256 932ebce85aebb89467cb9da0d82b136debed8b77f6a9f0b8e852336480db0adb. All entries share this exact revision. Status: partial source-derived relations, finite native evidence for pure seams, source inspection for effects. Domain: finite supplied inputs under the stated resource/arithmetic premises. Exclusions are inner source-derived families, not a complete behavioral complement. Reverse relations retain compatible antecedents but no complete inverse solver is implemented.
forward
Eligibility for (i,gd,gh,p,remembered) is gd>0, gh>=0 and p<0.05. Return eligible tuple with maximal gd then minimal i; None if no eligible row. remembered is carried but not ranked.
behavior-complement
Universe: unique-index result arrays with proposed None/tuple. Exclude ineligible selection, None despite an eligible row, or a row outranked by greater eligible dev gain or smaller tied index. NaN p is ineligible; negative/infinite-negative p remains allowed by legacy comparison.
reverse-reconstruction
Chosen row admits arbitrary rejected and lower-ranked extra rows. None admits every array of ineligible tuples, including empty. Original sample scores and journal validity cannot be reconstructed.
invariant
Return an input tuple unchanged. Unique indices give permutation invariance. Statistical interpretation assumes finite p in [0,1] but code does not enforce it.
cost
O(R) time, O(1) extra storage. Caller still sorts for reporting, so extraction adds a scan rather than a speed gain.
tie-stability
Distinct indices and fixed observations make all input permutations choose the same row. Finite oracle checks 60 combinations in forward/reverse orders; production indices derive from history-sorted proposals.
threshold-partition
dev<=0, validation<0 and p>=0.05/NaN each forbid adoption. Boundary p=0.05 is rejected. Compiled <= mutant fails rejection_thresholds_and_configuration_order.
composition
Actual caller sorts results for printing then invokes selector. With unique indices this equals former sorted-first eligible selection.
epoch-memory
Wave14 first modeling epoch, 2026-10-30: preserves rendering and tuning selection semantics. Two public pure tuning seams are now actual production-path calls. Six rendering tests and four tuning tests record field loss, tolerance/duplicate behavior, strict selection thresholds and domain limits. No bug repair or speed claim. Existing semantics retained include newline passthrough, first key/last MId collision, zip truncation and unchecked score arithmetic. Evidence is finite; no parser, teacher, filesystem corpus, journal-backed tune, optimizer training or GPU execution.
limitations
Partial syntax binds source bytes; macro bodies remain opaque and semantic AST is absent. Native/structural checks are not universal proofs or full-system coverage. No approximation error budget or automatic substitution is claimed. File evidence distinguishes executed pure paths and inspected effects.
This page in Markdown โ https://md.groupants.elamur.ai/free/en/mova/world/src/tune.rs.groupant/selection.groupant/