Iso induces order iso
ProvedOperadicStoneDuality.iso_induces_order_isoaether-catalogbridges
Formal statement of OperadicStoneDuality.iso_induces_order_iso from the Aether Catalog (Bridges). The mathematical content is given by the Lean statement below; a human-readable write-up is pending.
theorem OperadicStoneDuality.iso_induces_order_iso{N M : NeuralArchFG}
(h : UpperSet N.Module ≃o UpperSet M.Module) :
∃ f : N.Module ≃o M.Module,
∀ m, h (UpperSet.Ici m) = UpperSet.Ici (f m) := by sorry
Formalization Note Transplanted verbatim from the Aether Catalog source Bridges/OperadicStoneDuality.lean; the statement is byte-identical to the source declaration, elaborated with autoImplicit disabled in the platform environment.
Preamble
-- Thm stub generated from Bridges/OperadicStoneDuality.lean
import Mathlib
import Definitions.Def_Bridges_OperadicStoneDuality
/-!
# Operadic Stone Duality: Neural Heyting Semimodules and
Certified Architecture–Kripke Reconstruction
Bridge: connects Algebra (distributive lattices, Heyting algebras, Birkhoff duality)
to Machine Learning (neural architecture identifiability, operadic deep learning)
to Logic (intuitionistic Kripke semantics, prime filters, bounded morphisms).
## Overview
We prove that finitely generated acyclic neural architectures admit a canonical
intuitionistic semantics — a finite Heyting algebra of *monotone predicates*
(upper sets of the module poset) — from which the architecture can be
reconstructed up to isomorphism.
The upper set lattice `UpperSet N.Module` is a finite distributive lattice
and Heyting algebra. Its join-irreducible elements are the principal upper sets
`↑m = Ici m`, which correspond bijectively to modules. The module partial order
is recovered via `m₁ ≤ m₂ ↔ Ici m₁ ≤ Ici m₂` (in the UpperSet order).
An order isomorphism of upper-set lattices therefore induces an order isomorphism
of module posets, establishing the reconstruction theorem.
## Main Results
* `pred_distrib_lattice` — upper sets form a distributive lattice
* `pred_heyting` — upper sets form a Heyting algebra
* `pred_finite` — the lattice is finite
* `ici_orderEmbedding` — m ↦ Ici m is an order embedding
* `principalUpper_joinIrred` — principal upper sets are join-irreducible
* `joinIrred_iff_principal` — join-irreducibles = principal upper sets
* `soundness_completeness` — lattice order = Kripke semantic entailment
* `upperPredMap_contravariant` — contravariant functoriality
* `iso_induces_order_iso` — upper-set lattice iso ⟹ module order iso
* `semantics_determines_architecture` — the main reconstruction theorem
-/
open Set Function
noncomputable section
open OperadicStoneDuality
/-! ## Part I: Neural Architecture Foundations -/
attribute [instance] NeuralArchFG.instFintype NeuralArchFG.instDecEq
NeuralArchFG.instPartialOrder
/-! ## Part II: Upper Set Predicate Lattice
We use Mathlib's `UpperSet α` — the type of upward-closed subsets of a
preordered type. For a finite partial order, this is automatically:
- a distributive lattice (`DistribLattice`)
- a Heyting algebra (`HeytingAlgebra`)
- finite (`Finite`)
The ordering on `UpperSet α` is: `U ≤ V ↔ V.carrier ⊆ U.carrier` (reverse
inclusion). This means `Ici m₁ ≤ Ici m₂ ↔ m₁ ≤ m₂`, so the map `m ↦ Ici m`
is an order embedding. -/
/-! ## Part III: Principal Upper Sets and Order Embedding
The map `m ↦ Ici m` sends modules to their principal upper sets.
This is an order embedding, meaning the module partial order is
faithfully encoded in the upper set lattice. -/
/-
The principal upper set map is an order embedding.
This is the key structural fact: the module order is encoded
in the upper-set lattice order.
Bridge: Order theory (Birkhoff embedding) ↔ ML (architecture encoding).
-/
/-
The map `m ↦ Ici m` is injective.
Bridge: Algebra (separation).
-/
/-! ## Part IV: Join-Irreducibles
An element of a bounded lattice is join-irreducible if it's nonzero and
cannot be written as a non-trivial join. The join-irreducibles of
`UpperSet α` are exactly the principal upper sets `Ici m`. -/
/-
Principal upper sets are meet-irreducible (`InfIrred`) in the upper set lattice.
In Mathlib's `UpperSet`, `⊓ = ∪` and `⊔ = ∩`, so meet-irreducibility means
`Ici m` cannot be written as `A ∪ B` for strictly larger `A`, `B`.
Proof: if `Ici m = A ∪ B` then `m ∈ A` or `m ∈ B`. WLOG `m ∈ A`.
Since `A` is upper, `Ici m ⊆ A ⊆ A ∪ B = Ici m`, so `A = Ici m`.
Bridge: Algebra (meet-irreducible) ↔ ML (atomic module).
-/
/-
Meet-irreducible upper sets are exactly the principal upper sets.
Bridge: Algebra (classification of irreducibles) ↔ ML (module identification).
-/
/-
The meet-irreducibles biject with modules.
-/
/-! ## Part V: Soundness and Completeness -/
/-! ## Part VI: Architecture Morphisms and Contravariance -/
/-
**Theorem (Contravariant Functoriality):**
upperPredMap respects composition contravariantly.
-/
/-
upperPredMap preserves identity.
-/
/-! ## Part VII: Reconstruction Theorem -/
/-
**Key Lemma:** An order isomorphism of upper-set lattices preserves
meet-irreducibles.
-/
/-
**Key Lemma:** An order isomorphism of upper-set lattices induces
an order isomorphism of the module posets.
-/Formal statement
theorem OperadicStoneDuality.iso_induces_order_iso{N M : NeuralArchFG}
(h : UpperSet N.Module ≃o UpperSet M.Module) :
∃ f : N.Module ≃o M.Module,
∀ m, h (UpperSet.Ici m) = UpperSet.Ici (f m) := by sorrySource