Alphabeta Math
ExampleConstruction: AI-generatedVerification: AI-generatedPipeline-generatedjudge pass (gpt-6-luna)
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Fan-in and fan-out target different signal directions

Example

Compare the source's forward and atomless conditional backward recurrence coefficients for fan-in 64 and fan-out 128. The respective coefficients are 64σ2/2 and 128σ2/2. Compute their factors at the two initializer choices σ2=2/64 and σ2=2/128. For a preceding preactivation variance V the forward result is a multiple of V; for a downstream second moment G the conditional backward result is a multiple of G (and of its variance when centered). These coefficient comparisons do not assert a nonzero actual ReLU gradient model satisfying the full backward independence assumptions: at a nondegenerate hidden ReLU layer those assumptions force G=0.

Diagram

The top row represents the forward linear map and ReLU; the lower row shows the reverse-mode maps with Dℓ=diag⁡(1Zℓ,j>0). Each box denotes the full vector of the stated dimension, and the arrow labelled Wℓ denotes the 128×64 matrix, not a single scalar edge. Thus the forward fan-in is 64 and the backward fan-out is 128. The diagram illustrates these dimensions; the verification below uses the recurrence formulas directly.

X`¡12R64Z`2R128X`2R128W`½g`¡12R64h`2R128g`2R128WT`D`

Facts & Assumptions

Given: The stated dimensions, source recurrence coefficients, and their conditional interpretations for V and G.

[F1]

Forward mode uses weight variance 2/nℓ, and backward mode uses 2/n^ℓ. (Rectifier variance-preserving initialization)

[F2]

The forward preactivation variance factor is nℓσℓ2/2. (ReLU forward preactivation-variance recurrence)

[F3]

Under its stated conditional assumptions, the backward gradient second-moment factor is n^ℓσℓ2/2. (ReLU backward gradient-variance recurrence)

Verification

1.1givenF1F2F3algebra

With fan-in variance σ2=2/64=1/32, the forward factor is 12(64)(1/32)=1, so V is preserved. The backward factor is 12(128)(1/32)=2, so the gradient variance becomes 2G.

1.2givenF1F2F3algebra

With fan-out variance σ2=2/128=1/64, the backward factor is 12(128)(1/64)=1, so G is preserved. The forward factor is 12(64)(1/64)=1/2, so the next preactivation variance is V/2.

2.1step 1.1step 1.2algebra∎

Since the two widths differ, the required variances 1/32 and 1/64 differ, and no shared weight variance makes both recurrence coefficients equal to one.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

14 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources