---
title: "gparse-model-heads — Mova code model"
description: "Mova code model of gparse/src/model.rs: components with their labels, obernants and the model text."
url: "https://groupants.elamur.ai/free/en/mova/gparse/src/model.rs.groupant/heads.groupant/"
---
Elamur: [Mova](https://md.mova.elamur.ai/free/en/) · [Dictionary](https://md.dictionary.elamur.ai/free/en/) · [Research](https://md.research.elamur.ai/free/en/) · **Groupants**

This page as HTML: <https://groupants.elamur.ai/free/en/mova/gparse/src/model.rs.groupant/heads.groupant/>

# gparse-model-heads

The groupant beside gparse/src/model.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

- **kind:** groupant_fragment
- **status:** partial source bound model
- **source:** [`gparse/src/model.rs`](https://github.com/ElamurAI/mova-free/blob/main/gparse/src/model.rs)
- **source SHA-256:** `a380c81aa11433fd…`
- **covers:** rels gathers flattened rows/heads and evaluates bilinear+linear logits; tags projects gathered words. (bytes 13492–14585)
- **part of:** [gparse-model](https://md.groupants.elamur.ai/free/en/mova/gparse/src/model.rs.groupant/)
- **repository folder:** [`gparse/src/model.rs.groupant/heads.groupant`](https://github.com/ElamurAI/mova-free/tree/main/gparse/src/model.rs.groupant/heads.groupant)
- **repository commit:** `48f2e6f74146`

> **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 | [approximant](#approximant) |
| **ast** | partial | `../syntax.json` |
| **cost** | symbolic source derived | [invariant](#invariant) |
| **descriptant** | present | [descriptant](#descriptant) |
| **invariant** | partial | [invariant](#invariant) |
| **tests** | partial bounded checks | [tests](#tests) |

## 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 |
|---|---|---|---|
| `behavior_complement` | behavior_complement | partial source derived model | [behavior_complement](#behavior-complement) |
| `reverse_reconstruction` | reverse_reconstruction | partial source derived model | [reverse_reconstruction](#reverse-reconstruction) |
| `gather_coordinates` | gather_coordinates | partial source derived model | [gather_coordinates](#gather-coordinates) |
| `relation_equation` | relation_equation | partial source derived model | [relation_equation](#relation-equation) |
| `tag_projection` | tag_projection | partial source derived model | [tag_projection](#tag-projection) |
| `dropout_effect` | dropout_effect | partial source derived model | [dropout_effect](#dropout-effect) |

## Model text


### descriptant

#### Source-bound description

rels gathers flattened rows/heads and evaluates bilinear+linear logits; tags projects gathered words.

##### Epoch 2026-10-31

Prospective model was recorded before edits. Added tests use bounded synthetic fixtures; production algorithms unchanged. Coverage is partial, not proof. No trained corpus, live teacher or throughput experiment was executed.

### approximant

#### Forward approximant

rels gathers flattened rows/heads and evaluates bilinear+linear logits; tags projects gathered words.

Domain: valid source-specific dimensions and resources stated in invariant; no automatic production substitution. Numerical observations are finite checks, not an end-to-end relative bound. Executable Rust surrogate: unsupported gap; existing exact tests are not surrogates. Future metric-specific caps may be 1/2/5/10%, never above 10%.

### invariant

#### Invariant and cost

rows and heads refer to same flattened B*T coordinate frame; classes index vocab ordering.

##### Cost

O(N*d*r+N*r²*K+N*r*K+N*d*(X+U)) time for relation projections/bilinear/linear and tag projections, plus O(N*d) gathers; K/X/U class counts independent.

Allocation, finite arithmetic, platform dimensions and external failures are not universally certified.

### models

## Typed model catalog

### behavior_complement

A local head index without k*T offset does not represent later sentence root.

### reverse_reconstruction

Argmax tags/relations admit cones of logits, not unique hidden state.

### gather_coordinates

flat out.h shape B*T,d; rows and heads index first axis.

### relation_equation

dd*rel_w reshape N,L,r then elementwise hh and sum plus concat-linear.

### tag_projection

XPOS/UPOS linear classifiers share gathered encoder z.

### dropout_effect

Relation projections use train dropout; tags function itself has no train/dropout parameter.

### tests

#### Checks and limits

Selected parent checks relevant to this fragment (not exhaustive): model::tests::embeddings_get_gradients; model::fused::which_ops_backprop; activation_configuration_roundtrips; CUDA gradient/GELU/synthetic optimizer checks; CPU prediction

Selected commands run twice debug and twice release; exact receipts and source-to-test mapping retained in wave30 gparse external run directory. Tests do not certify allocation failure, all NaN/overflow inputs, trained-model accuracy or corpus timing. Compiled mutants cover selected boundaries; per-source coverage is recorded in the external map, not presumed.