ReLUNetwork
structure
A feedforward ReLU neural network from the real line to the real line with hidden layers of width k, specified by rational affine input and output layers and an arbitrary sequence of width-k hidden layers.
Statement
structure ReLUNetwork (k : ℕ) where in_weight : Fin k → ℚ in_bias : Fin k → ℚ layers : List (ReLUHiddenLayer k) out_weight : Fin k → ℚ out_bias : ℚ
- Defined in
- arXiv:2610.06092, Section 2.1
- Verified
- 20 Oct 2026
- Axioms
- none listed
- Built on 1
- 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 11
- 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_network_has_unit_interval_zeroA feedforward ReLU neural network
netof width $k$ has a zero in the unit interval $[0, 1]$ if there exists some real number… - 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…
- 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_compact_networkThe width-three network obtained by concatenating a seven-layer block for each integer weight, followed by the final target-checking layer…
- relu_network3_encode_injectiveSanity check of the encoding: distinct width-3 ReLU networks have distinct Bool encodings.
- 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…
- relu_network_eval_continuousFor every natural number $k$ and every feedforward ReLU neural network
netof width $k$ with rational weights and biases, the evaluation… - relu_network3_encodeEncoding of a width-3 ReLU network as a Bool word, of length linear in the total bit length of its rational entries: input weights, input…
- 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
A structure ReLUNetwork parameterized by a natural number $k$, with fields:
- in_weight: a function from Fin k to ℚ (a rational vector of dimension $k$),
- in_bias: a function from Fin k to ℚ (a rational vector of dimension $k$),
- layers: a list of ReLUHiddenLayer k,
- out_weight: a function from Fin k to ℚ (a rational vector of dimension $k$),
- out_bias: a rational number (ℚ), with a default value given by by sorry.
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
/-- A feedforward ReLU neural network from ℝ to ℝ with hidden layer width k and rational weights and biases. -/ structure ReLUNetwork (k : ℕ) where in_weight : Fin k → ℚ in_bias : Fin k → ℚ layers : List (ReLUHiddenLayer k) out_weight : Fin k → ℚ out_bias : ℚ