LeanProofs.Rosetta.ObstructionToDescent
DefinitionGrade4_LeanProofs_Rosetta_ObstructionToDescentclifford-algebraleanproofs-grade4rosetta
Foundational definitions and structural lemmas from LeanProofs.Rosetta.ObstructionToDescent, required by the selected LeanProofs theorem graph.
Definition code
import Mathlib.Tactic
namespace LeanProofsGrade4P2M
/-!
# Obstructions to descent through an observational quotient
This module isolates a generic finite-logical pattern used throughout the
Rosetta research boundary. A full-state update `step : State → State`
descends through an observation `observe : State → Coarse` when one autonomous
coarse update reproduces every observed full-state step.
The fiber criterion says that descent is equivalent to respecting the fibers
of `observe`. A typed obstruction is a pair of observationally identical
states whose next observations differ. Such a witness rules out every
autonomous coarse update, independently of how that update is proposed.
An additional memory coordinate repairs this particular obstruction only if
it distinguishes the witness states. Retaining the complete state is always
sufficient, while a constant memory coordinate adds no expressive power.
These are logical factorization results. They do not identify a physical
state, dynamics, coarse graining, stochastic memory kernel, or minimal model.
-/
namespace Rosetta.ObstructionToDescent
universe u v w z
variable {State : Type u} {Coarse : Type v} {Memory : Type w} {Outcome : Type z}
/-! ## Generic observational factorization -/
/-- An outcome is a function of the observation alone. -/
def FactorsThrough (observe : State → Coarse) (outcome : State → Outcome) : Prop :=
∃ coarseOutcome : Coarse → Outcome,
∀ state, outcome state = coarseOutcome (observe state)
/-- Equal observations determine equal outcomes. -/
def Determines (observe : State → Coarse) (outcome : State → Outcome) : Prop :=
∀ ⦃left right⦄, observe left = observe right → outcome left = outcome right
/-- Equal observations with unequal outcomes obstruct every factorization. -/
structure DeterminationObstruction
(observe : State → Coarse) (outcome : State → Outcome) where
left : State
right : State
same_observation : observe left = observe right
different_outcome : outcome left ≠ outcome right
/-- Observation augmented by an explicitly supplied memory coordinate. -/
def augment (observe : State → Coarse) (remember : State → Memory) :
State → Coarse × Memory :=
fun state => (observe state, remember state)
/-- A memory coordinate is sufficient for a selected outcome on every fiber. -/
def IsSufficientFor (observe : State → Coarse) (outcome : State → Outcome)
(remember : State → Memory) : Prop :=
Determines (augment observe remember) outcome
/-! ## Autonomous descent -/
end Rosetta.ObstructionToDescent
end LeanProofsGrade4P2M
Source