Marked Reverse Execution Restores the Start of an Interior Trace
ProvedMonodromy.CoupledContinuation.Program.reverseReturn_of_firstReturnLet d>0 and let F,R,B be programs of degree d such that R.advance is a left inverse of F.advance on every coupled state. Suppose F.step(s)=some u, a finite possibly empty chain of F.steps leads from u to t with B.haltSignal≠1 at every departure state, and B.haltSignal(s)=1. Then reverseReturn R B from t is defined and equals s: first take R.advance once, then evaluate R’s words with B’s stopping bits. No separate boundary-mark hypothesis on t is required.
Statement dependencies: Monodromy.CoupledContinuation.Memory, Monodromy.CoupledContinuation.Point, Monodromy.CoupledContinuation.Program, Monodromy.CoupledContinuation.Program.advance, Monodromy.CoupledContinuation.Program.haltSignal, Monodromy.CoupledContinuation.Program.reverseReturn, Monodromy.CoupledContinuation.Program.step.
Original proof dependencies: Monodromy.CoupledContinuation.Memory, Monodromy.CoupledContinuation.Point, Monodromy.CoupledContinuation.Program, Monodromy.CoupledContinuation.Program.advance, Monodromy.CoupledContinuation.Program.boundaryInverse, Monodromy.CoupledContinuation.Program.eval, Monodromy.CoupledContinuation.Program.haltSignal, Monodromy.CoupledContinuation.Program.reverseReturn, Monodromy.CoupledContinuation.Program.step.
Statement elaboration dependencies: Monodromy.CoupledContinuation.Memory, Monodromy.CoupledContinuation.Point, Monodromy.CoupledContinuation.Program, Monodromy.CoupledContinuation.Program.advance, Monodromy.CoupledContinuation.Program.haltSignal, Monodromy.CoupledContinuation.Program.reverseReturn, Monodromy.CoupledContinuation.Program.step.
Proof elaboration dependencies: Monodromy.CoupledContinuation.Memory, Monodromy.CoupledContinuation.Point, Monodromy.CoupledContinuation.Program, Monodromy.CoupledContinuation.Program.advance, Monodromy.CoupledContinuation.Program.boundaryInverse, Monodromy.CoupledContinuation.Program.eval, Monodromy.CoupledContinuation.Program.haltSignal, Monodromy.CoupledContinuation.Program.reverseReturn, Monodromy.CoupledContinuation.Program.step.
module
public import Definitions.Def_Monodromy_Continuation
public import Lean.Elab.Tactic.Omega
public import Mathlib.Analysis.Analytic.Constructions
public import Mathlib.Analysis.Analytic.IsolatedZeros
public import Mathlib.Analysis.Analytic.Linear
public import Mathlib.Analysis.Calculus.DSlope
public import Mathlib.Analysis.Calculus.Deriv.Add
public import Mathlib.Analysis.Calculus.Deriv.Comp
public import Mathlib.Analysis.Calculus.Deriv.Mul
public import Mathlib.Analysis.Calculus.Deriv.Pow
public import Mathlib.Analysis.Complex.Basic
public import Mathlib.Analysis.Complex.CoveringMap
public import Mathlib.Analysis.Real.Pi.Bounds
public import Mathlib.Analysis.SpecialFunctions.Complex.Analytic
public import Mathlib.Analysis.SpecialFunctions.Complex.Circle
public import Mathlib.Computability.StateTransition
public import Mathlib.Data.Fin.Basic
public import Mathlib.Data.Fintype.Card
public import Mathlib.Logic.Equiv.Fin.Rotate
public import Mathlib.Logic.Function.Conjugate
public import Mathlib.RingTheory.RootsOfUnity.Complex
public import Mathlib.Tactic.FieldSimp
public import Mathlib.Tactic.FunProp
public import Mathlib.Tactic.Linarith
public import Mathlib.Tactic.Ring
public import Mathlib.Topology.Homotopy.Lifting
@[expose] public section
open Monodromy
open Monodromy.CoupledContinuation
open Monodromy.CoupledContinuation.Program
variable {degree : ℕ} [NeZero degree]
theorem Monodromy.CoupledContinuation.Program.reverseReturn_of_firstReturn (forward backward boundary : Program degree)
(inverse : Function.LeftInverse backward.advance forward.advance)
{first after last : Memory} (initial : forward.step first = some after)
(interior : Relation.ReflTransGen
(fun left right => forward.step left = some right ∧ boundary.haltSignal left ≠ 1)
after last)
(marked : boundary.haltSignal first = 1) :
reverseReturn backward boundary last = Part.some first := by sorry
Read-back
What the Lean code literally says, in plain math · OpenAI Codex (independent agent)
Let be a positive natural number and let be arbitrary geometric word programs with table entries and stop flags. Their memories are all triples with , , and . Write for the total advances of , each executing the word selected by the root ; for a program , let its stopping signal be , where and is the removable divided difference of at , divided by . A step of has no successor when that signal is , and otherwise returns its total advance. Assume for every memory . For arbitrary memories , assume one successful -step sends to , that reaches by finitely many successful -steps every one of whose source memories has -signal different from , and that the -signal at is . Then first applying to and subsequently evaluating the program with 's words and 's stop flags is a defined partial computation whose result is exactly . The finite interior sequence may be empty; there is no assumption that is marked, no right-inverse hypothesis, and no requirement that the three programs have the same stop flags. In particular, if the hypotheses about -stepping and the initial mark conflict for a particular choice of programs, that instance has no qualifying states.
Confirmed by the mission captain (proposal self-audit).