Purpose & architecture
The shared implementation embeds integer-like inputs using a sinusoidal table, combines them with positional weights, applies dense layers and GELU, then tries to produce vocabulary probabilities. Its Train method computes a cross-entropy-like loss without updating weights. The implementation is a local numerical scaffold.
Static review finds a blocking dimension error: LinearLayer returns HiddenDim values, while Forward slices that result to vocabSize. The named constructors use vocabularies larger than the hidden dimension, so that slice is out of bounds. The familiar model labels do not supply their official architectures or pretrained weights.
This source generation
This record follows the claude/improve-math-logic-ehIW1 branch of Janady13/complete-agi-core at commit 33aa84f2aa61cedbe010240ada7f4efbdfe795a7. The snapshot identifies the source inspected in this audit. It is not a statement about which revision is currently running on a device.
| Field | Source record |
|---|---|
| Revision | 33aa84f2aa61cedbe010240ada7f4efbdfe795a7 |
| Branch | claude/improve-math-logic-ehIW1 |
| Recorded generation | Generator Core · autoregressive |
Recovered implementation structure
These paths locate the components and entry points represented by this build. The source register contains their inspected revisions.
- generator_core/pkg/generators/autoregressive/autoregressive.go
Behavior & verification limits
Static review finds a blocking dimension error: LinearLayer returns HiddenDim values, while Forward slices that result to vocabSize. The named constructors use vocabularies larger than the hidden dimension, so that slice is out of bounds. The familiar model labels do not supply their official architectures or pretrained weights.
Original project documentation
The following documentation is preserved as a dated source account. Claims in it retain the source’s wording and must be read alongside the implementation notes above.
Package autoregressive implements autoregressive generators (1-10) P(x) = Π_t P(x_t | x_1, ..., x_{t-1}) ═══════════════════════════════════════════════════════════════════════════════ AUTOREGRESSIVE GENERATOR BASE ═══════════════════════════════════════════════════════════════════════════════ AutoregressiveGenerator implements autoregressive sequence generation NewAutoregressiveGenerator creates a new autoregressive generator Initialize initializes the autoregressive generator
Equations in this build
1 linked mathematical records. Each opens its notation, interpretation and source.