MLP: first-layer mean vanishes (eq. 4.6)
ProvedDeepLearningTheory.mlp_first_layer_mean_zeroFor an MLP at initialization (eqs. 4.3–4.4), every input and every first-layer neuron ,
import Mathlib import Definitions.Def_DLT_DeepLinearNetwork import Definitions.Def_DLT_MLPInit open MeasureTheory ProbabilityTheory open scoped NNReal
namespace DeepLearningTheory
theorem mlp_first_layer_mean_zero {Ω : Type*} [MeasurableSpace Ω] (P : Measure Ω)
[IsProbabilityMeasure P] (n : ℕ → ℕ) (Cb CW : ℕ → ℝ≥0) (σ : ℝ → ℝ)
(b : Ω → ℕ → ℕ → ℝ) (W : Ω → ℕ → ℕ → ℕ → ℝ) (hinit : IsMLPInit P n Cb CW b W)
(x : ℕ → ℝ) (i : ℕ) (hi : i < n 1) :
∫ ω, mlpPreact n σ (b ω) (W ω) x 1 i ∂P = 0 := by sorry
end DeepLearningTheoryRead-back
What the Lean code literally says, in plain math · Aristotle (Harmonic) - non-blind, same agent as the drafter
Non-blind read-back. This read-back was written by the same agent that drafted the Lean statement, with full knowledge of the source text and of the intended meaning. It is not independent testimony and must not be mistaken for a blind audit; reviewers should compare the Lean code against the source themselves.
Let be a probability space, widths, , any function, and random bias/weight fields satisfying IsMLPInit (all biases and weights with , , mutually independent; , ). Write for the MLP preactivation at and for the first-layer metric of the listed inputs. Then for every input and every : , where .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.