AFTD Auto-Formalizing Theoretical Domains

op75_compact_network

def

The width-three network obtained by concatenating a seven-layer block for each integer weight, followed by the final target-checking layer and an all-one output.

Statement

def op75_compact_network (as : List ℕ) (t : ℕ) : ReLUNetwork 3
Source
OP-75, arXiv:2610.06092; explicit seven-layer single-pass subset-sum construction.
Verified
15 Nov 2026
Axioms
none listed
Built on 4
  • ReLUNetworkA feedforward ReLU neural network from the real line to the real line with hidden layers of width k, specified by rational affine input…
  • op75_compact_blockSeven width-three hidden layers for one nonnegative integer weight in the one-pass subset-sum construction.
  • op75_compact_finalA width-three layer that forwards the orbit coordinate and computes the two positive parts of the accumulated sum minus the target.
  • ReLUHiddenLayerA hidden layer of width k consists of a k-by-k rational weight matrix and a rational bias vector of length k.
Used by 7

Lean source view module on GitHub

def op75_compact_network (as : List ℕ) (t : ℕ) : ReLUNetwork 3 where
  in_weight := ![1,0,0]
  in_bias := ![0,0,0]
  layers := as.flatMap op75_compact_block ++ [op75_compact_final t]
  out_weight := ![1,1,1]
  out_bias := 0
See it in the graphDiscuss this resultChallenge it