Purpose & architecture
The shared implementation chains affine coupling transformations and permutations, maps inputs into a latent space, combines a standard-normal log density with accumulated log-Jacobian terms and can generate samples by transforming random latent vectors. Train returns a negative log-likelihood-style objective.
Invertibility, transformation inverses and Jacobian signs require numerical verification. The names select this shared local family implementation; they are not imported official models. The inspected Train method only returns a loss.
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 · flow |
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/flow/flow.go
Behavior & verification limits
Invertibility, transformation inverses and Jacobian signs require numerical verification. The names select this shared local family implementation; they are not imported official models. The inspected Train method only returns a loss.
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 flow implements normalizing flow generators (56-65) p(x) = p(z) |det ∂f⁻¹/∂x| where z = f⁻¹(x) ═══════════════════════════════════════════════════════════════════════════════ FLOW GENERATOR BASE ═══════════════════════════════════════════════════════════════════════════════ FlowGenerator implements normalizing flow models NewFlowGenerator creates a new flow generator Initialize initializes the flow generator
Equations in this build
3 linked mathematical records. Each opens its notation, interpretation and source.
flow · p(x)
This relation defines p(x) in the source section “flow · p(x)”. Uses a derivative or gradient to describe a rate of change or a direction of variation. Its interpretation depends…
func (g *FlowGenerator) affineTransform(x []float64, flowIdx int, inverse bool) ([]float64, float64) · Forward: y2
This relation defines Forward: y2 in the source section “func (g *FlowGenerator) affineTransform(x []float64, flowIdx int, inverse bool) ([]float64, float64) · Forward: y2”.…
func (g *FlowGenerator) affineTransform(x []float64, flowIdx int, inverse bool) ([]float64, float64) · Inverse: x2
This relation defines Inverse: x2 in the source section “func (g *FlowGenerator) affineTransform(x []float64, flowIdx int, inverse bool) ([]float64, float64) · Inverse: x2”.…