Coordinate tensors and matrix slot actions
DefinitionKLSModular_FunctionalInequalities_Analysis_Tensor_SlotActionfunctional-inequalitiesklskls-conjecturekls-modularlog-concavemodule-context
An order- coordinate tensor is represented as a function . inner T S is its Euclidean coordinate inner product ; actOn H s T applies a matrix to the -th tensor input, and actSum H T sums this action over all slots. These definitions use Function.update to preserve every tensor index except the selected one.
Original module imports
import Mathlib
Original declaration excerpt
Shown in the original module context.
def inner (T S : (Fin r → Fin n) → ℝ) : ℝ := ∑ J, T J * S J
Module navigation
Theorems in source order:
Project result: The KLS conjecture.
Original module · FunctionalInequalities.Analysis.Tensor.SlotAction · Frozen source 408915f.
Definition code
import Mathlib
noncomputable section
open Matrix
namespace FunctionalInequalities
namespace Analysis
namespace CoordinateTensor
variable {n r : ℕ}
/-- The Euclidean inner product of two order-`r` coordinate tensors on `ℝⁿ`. -/
def inner (T S : (Fin r → Fin n) → ℝ) : ℝ := ∑ J, T J * S J
/-- `H^{(s)}`: the matrix `H` acting on the `s`-th input of an order-`r` tensor (BKL arXiv:2610.05474v1, p. 23). -/
def actOn (H : Matrix (Fin n) (Fin n) ℝ) (s : Fin r) (T : (Fin r → Fin n) → ℝ) : (Fin r → Fin n) → ℝ :=
fun J => ∑ k, H (J s) k * T (Function.update J s k)
/-- `Σ_H = Σ_s H^{(s)}`. -/
def actSum (H : Matrix (Fin n) (Fin n) ℝ) (T : (Fin r → Fin n) → ℝ) : (Fin r → Fin n) → ℝ :=
fun J => ∑ s, actOn H s T J
end CoordinateTensor
end Analysis
end FunctionalInequalities
Source
Lean source: https://github.com/human-ai-mathematics/kls-lean/blob/408915f/formalization/FunctionalInequalities/Analysis/Tensor/SlotAction.lean#L19-L28 (repository human-ai-mathematics/kls-lean, commit 408915f, module `FunctionalInequalities.Analysis.Tensor.SlotAction`, definition bundle `KLSModular_FunctionalInequalities_Analysis_Tensor_SlotAction`; the repository is private at upload time).