Eq. (123): sum-over-paths formula for a ReLU network
ProvedLesHouchesWidth.output_eq_sum_over_pathsFor every configuration of weights, every input and every output neuron ,
where is the set of paths through the network with and .
This identity expresses a ReLU network with zero biases as a polynomial in its weights whose coefficients are the activation indicators. All the moment computations of Lecture 5 start from it.
Formalization Note The statement is deterministic, with no probability involved. The notes index weights along a path as ; here the row/column orientation of the network recursion is used.
import Mathlib import Definitions.Def_LesHouchesWidth_ReLUNet
namespace LesHouchesWidth
open MeasureTheory ProbabilityTheory
theorem output_eq_sum_over_paths {n : ℕ → ℕ} {L : ℕ} (ω : Weights n L)
(x : Fin (n 0) → ℝ) (q : Fin (n (L + 1))) :
output ω x q =
∑ p : Fin (n 0), x p *
∑ γ ∈ pathsFromTo n L p q, pathWeight ω γ * pathActivation ω x γ := by sorry
end LesHouchesWidthRead-back
What the Lean code literally says, in plain math · Aristotle (Harmonic) - non-blind, same agent as drafter
Non-blind read-back. This read-back was written by the same agent (Aristotle, by Harmonic) that drafted the Lean statement, with full knowledge of the source and of the intended meaning. It is not independent testimony and should not be treated as a blind audit; an independent read-back is still recommended before launch.
Let be any width sequence and any depth ( is allowed). Let be any real weight configuration (no probability is involved), let be any input, and let . The theorem asserts the exact identity
Here is the zero-bias ReLU recursion , . The weight is , and is the set of all index sequences with , , . There are no hypotheses. If some width is , both sides are empty sums equal to . The indicator uses the strict inequality , so a neuron whose preactivation is exactly contributes .
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.