relu_network_has_unit_interval_zero
def
A feedforward ReLU neural network net of width $k$ has a zero in the unit interval $[0, 1]$ if there exists some real number $x \in [0, 1]$ such that evalNetwork net x = 0.
Statement
def relu_network_has_unit_interval_zero {k : ℕ} (net : ReLUNetwork k) : Prop
- Defined in
- arXiv:2610.06092 Corollary 5.12
- Verified
- 25 Oct 2026
- Axioms
- none listed
- Built on 2
- evalNetworkEvaluation of a feedforward ReLU neural network of hidden layer width $k$ on a real input $x$, which forms an initial width-$k$ vector by…
- ReLUNetworkA feedforward ReLU neural network from the real line to the real line with hidden layers of width k, specified by rational affine input…
- Used by 8
- op75_compact_network_subset_sumFor any finite list of nonnegative integer weights and nonnegative integer target, the explicitly constructed width-three ReLU network has…
- op75_invalid_input_fallback_nonmemberThe encoding of the empty-weight target-one network is a no-instance of the width-three unit-interval zero language.
- relu_network_zero_of_ivtIf a feedforward ReLU neural network of width $k$ evaluates to a non-positive value at 0 and a non-negative value at 1, then by the…
- relu_network_width4_liftingWidth-4 lifting theorem (Theorem 5.8 of arXiv:2610.06092): For every 3-CNF formula $f$ with variable indices in $\mathbb{N}$, there exists…
- relu_width3_unit_interval_zero_languageThe verification language at width 3 on the unit interval, for the fixed encoding relu_network3_encode: the set of Bool words w such that…
- op75_subset_compiler_semanticsAny total binary function that maps canonical subset-sum instances to their specified networks and maps malformed inputs outside the zero…
- relu_width3_language_mem_iffSanity check of the language: the encoding of a width-3 ReLU network lies in the verification language exactly when the network has a zero…
- op_75_relu_width3_liftingOpen Problem OP-75 (arXiv:2610.06092, Remark 5.15): For every 3-CNF formula $f$ with variable indices in $\mathbb{N}$, there exists a…
Read back from the Lean
For a fixed (implicit) natural number k and a ReLU neural network net of width k with rational weights and biases, this defines the proposition that there exists a real number x in the closed unit interval [0, 1] such that evalNetwork net x = 0.
Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.
Lean source view module on GitHub
/-- Predicate asserting that a feedforward ReLU network has a root in the unit interval [0,1]. -/ def relu_network_has_unit_interval_zero {k : ℕ} (net : ReLUNetwork k) : Prop := ∃ x ∈ Set.Icc (0 : ℝ) 1, evalNetwork net x = 0