QPBT definitions: Test / MagicSquareTheorems / Rigidity / Reflections
DefinitionQPBT_QPBT_Test_MagicSquareTheorems_Rigidity_Reflectionsqpbt
This bundle defines state-dependent closeness, binary measurement observables, and the operator identities used in the rigidity estimates. It records linearity, homogeneity, norm preservation under isometries, and projectivity of joint local effects.
Principal definitions:
MIPStarRE.QPBT.MagicSquareRigidity.NormCloseOn: defines closeness of two operators within a state-dependent tolerance.MIPStarRE.QPBT.MagicSquareRigidity.signObs: defines the plus-or-minus-one observable associated with a binary measurement and answer relabeling.MIPStarRE.QPBT.MagicSquareRigidity.heteroKron_sum_left: states that finite sums pass through the first tensor factor.MIPStarRE.QPBT.MagicSquareRigidity.heteroKron_sum_right: states that finite sums pass through the second tensor factor.MIPStarRE.QPBT.MagicSquareRigidity.applyOperatorToState_smul: states that operator action on a state is homogeneous.MIPStarRE.QPBT.MagicSquareRigidity.norm_applyOperatorToState_of_isometry: states that an isometric operator preserves the norm of every state.MIPStarRE.QPBT.MagicSquareRigidity.norm_applyOperatorToState_isometry_mul: states that left multiplication by an isometry preserves state-dependent norm.MIPStarRE.QPBT.MagicSquareRigidity.jointMeasurement: states that joint effects of two projective local measurements form a projective measurement on the composite space.
Formalization Credit. This Lean declaration is part of the Quantum Pauli basis test formalization by Ruixuan Deng. The source code was generated by AI agents; Ruixuan Deng is the responsible maintainer.
Mathematical Source. The mathematical source is Ji, Natarajan, Vidick, Wright, and Yuen, MIP*=RE, arXiv:2001.04383v3.
Definition code
-- Generated from QPBT/Test/MagicSquareTheorems/Rigidity/Reflections.lean by skeleton subtraction.
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_AxisParallelLine
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_DiagonalLine
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_LinePolynomials
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_LowDegreePolynomial
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_OpFamily
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_ParametersBase
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_ParametersFiniteAnswers
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_QuantumState
import Definitions.Def_LIDT_MIPStarRE_LDT_CommutativityPoints_Defs
import Definitions.Def_LIDT_MIPStarRE_LDT_ExpansionHypercubeGraph_Defs_Core
import Definitions.Def_LIDT_MIPStarRE_LDT_GlobalVariance_Defs_Core
import Definitions.Def_LIDT_MIPStarRE_LDT_MakingMeasurementsProjective_Defs
import Definitions.Def_LIDT_MIPStarRE_LDT_Pasting_Defs_Families
import Definitions.Def_LIDT_MIPStarRE_LDT_Pasting_Defs_Interpolation
import Definitions.Def_LIDT_MIPStarRE_LDT_Tactic_AvgCongr
import Definitions.Def_LIDT_MIPStarRE_LDT_Test_StrategyCore
import Definitions.Def_LIDT_MIPStarRE_Quantum_FiniteMatrix_Basic
import Definitions.Def_LIDT_MIPStarRE_Quantum_Measurement
import Definitions.Def_QPBT_QPBT_Games_Defs
import Definitions.Def_QPBT_QPBT_Games_DistanceTheorems_Support
import Definitions.Def_QPBT_QPBT_Test_MagicSquareTheorems_Rigidity_GroundSlice
import Definitions.Def_QPBT_QPBT_Test_MagicSquareTheorems_Rigidity_Relations
namespace LIDT.MIPStarRE.LDT
end LIDT.MIPStarRE.LDT
namespace LIDT.MIPStarRE.Quantum
end LIDT.MIPStarRE.Quantum
namespace MIPStarRE.QPBT.MagicSquareRigidity
end MIPStarRE.QPBT.MagicSquareRigidity
namespace MIPStarRE.QPBT.MagicSquareRigidity.NormCloseOn
end MIPStarRE.QPBT.MagicSquareRigidity.NormCloseOn
/-!
# Reflections and the state-dependent metric on operators
The self-testing argument for the Magic Square game manipulates `±1`-valued
observables rather than positive operator valued measures. This file records
the algebra of such observables and the estimates that convert Born masses into
operator bounds.
Postprocessing a projective measurement along a binary relabelling of its
answers produces a *reflection*, that is, a self-adjoint involution; two
reflections coming from one and the same measurement commute, and their product
is the reflection attached to the sum of the two relabellings. The reflections
of the source are the binary observables of the rest of the development, so this
file uses `MIPStarRE.QPBT.IsBinaryObservable` and its algebra, defined in
`MIPStarRE/QPBT/Combining/Linearity/Defs.lean`, rather than a second predicate of
its own. Two operators are compared through the norm of the difference of their
actions on a fixed state, the relation `NormCloseOn` below. That relation is the
unsquared Euclidean distance between `M ψ` and `N ψ`; it is *not* the
state-dependent distance `≈_δ` of the source, which bounds the squared quantity
`⟨ψ|(M - N)ᴴ (M - N)|ψ⟩` and is formalized in this development by `opDistSq`.
The comparison is symmetric, satisfies the triangle inequality, and is unchanged
by multiplication on the left by an isometric operator; `NormCloseOn.opDistSq_le`
converts it into the source's relation. Finally, a real spectral combination of
the joint effects of the two players has state-dependent norm controlled by the
Born mass of the answer pairs carrying a nonzero coefficient; this is the step
that turns each rejection mass of the value-to-parity layer into an operator
estimate.
## References
The statements supported here are the operator relations in `thm:ms-rigidity`,
blueprint `blueprint/src/chapter/ch13_qpbt_test.tex:224-253`, from
`references/qpbt-paper/08_classical_and_quantum_low_degree_tests.tex:612-652`,
proved in Coladangelo--Stark, arXiv:1709.09267v2, Theorem 6.9.
-/
@[expose] public section
open scoped BigOperators Matrix MatrixOrder ComplexOrder
namespace MIPStarRE.QPBT.MagicSquareRigidity
open LIDT.MIPStarRE.LDT LIDT.MIPStarRE.Quantum
noncomputable section
/-! ## Rectangular tensor placement -/
section Kron
/-- Formalization-only: finite sums pass through the first factor. -/
theorem heteroKron_sum_left {ιA ιB γ : Type} [Fintype γ]
(M : γ → _root_.LIDT.MIPStarRE.Quantum.Op ιA) (N : _root_.LIDT.MIPStarRE.Quantum.Op ιB) :
_root_.MIPStarRE.QPBT.heteroKron (∑ c, M c) N = ∑ c, _root_.MIPStarRE.QPBT.heteroKron (M c) N := by
ext p q
simp [_root_.MIPStarRE.QPBT.heteroKron, Matrix.kronecker, Matrix.sum_apply, Finset.sum_mul]
/-- Formalization-only: finite sums pass through the second factor. -/
theorem heteroKron_sum_right {ιA ιB γ : Type} [Fintype γ]
(M : _root_.LIDT.MIPStarRE.Quantum.Op ιA) (N : γ → _root_.LIDT.MIPStarRE.Quantum.Op ιB) :
_root_.MIPStarRE.QPBT.heteroKron M (∑ c, N c) = ∑ c, _root_.MIPStarRE.QPBT.heteroKron M (N c) := by
ext p q
simp [_root_.MIPStarRE.QPBT.heteroKron, Matrix.kronecker, Matrix.sum_apply, Finset.mul_sum]
end Kron
/-! ## The state-dependent norm -/
section StateNorm
variable {ι : Type} [Fintype ι] [DecidableEq ι]
/-- Formalization-only: the action of an operator on a state is homogeneous. -/
theorem applyOperatorToState_smul (c : ℂ) (M : _root_.LIDT.MIPStarRE.Quantum.Op ι) (ψ : EuclideanSpace ℂ ι) :
_root_.MIPStarRE.QPBT.applyOperatorToState (c • M) ψ = c • _root_.MIPStarRE.QPBT.applyOperatorToState M ψ := by
unfold _root_.MIPStarRE.QPBT.applyOperatorToState
simp only [map_smul, LinearMap.smul_apply]
/-- An isometric operator preserves the norm of every state. -/
theorem norm_applyOperatorToState_of_isometry {U : _root_.LIDT.MIPStarRE.Quantum.Op ι} (hU : Uᴴ * U = 1)
(ψ : EuclideanSpace ℂ ι) : ‖_root_.MIPStarRE.QPBT.applyOperatorToState U ψ‖ = ‖ψ‖ := by
have h : ‖_root_.MIPStarRE.QPBT.applyOperatorToState U ψ‖ ^ 2 = ‖ψ‖ ^ 2 := by
rw [_root_.MIPStarRE.QPBT.MagicSquareRigidity.norm_applyOperatorToState_sq, hU, @norm_sq_eq_re_inner ℂ,
_root_.MIPStarRE.QPBT.MagicSquareRigidity.applyOperatorToState_one]
rfl
have h1 : (0 : ℝ) ≤ ‖_root_.MIPStarRE.QPBT.applyOperatorToState U ψ‖ := norm_nonneg _
have h2 : (0 : ℝ) ≤ ‖ψ‖ := norm_nonneg _
nlinarith [h, h1, h2]
/-- Multiplying on the left by an isometric operator does not change the
state-dependent norm. -/
theorem norm_applyOperatorToState_isometry_mul {U : _root_.LIDT.MIPStarRE.Quantum.Op ι} (hU : Uᴴ * U = 1)
(M : _root_.LIDT.MIPStarRE.Quantum.Op ι) (ψ : EuclideanSpace ℂ ι) :
‖_root_.MIPStarRE.QPBT.applyOperatorToState (U * M) ψ‖ = ‖_root_.MIPStarRE.QPBT.applyOperatorToState M ψ‖ := by
rw [_root_.MIPStarRE.QPBT.DistanceCalculus.applyOperatorToState_mul,
_root_.MIPStarRE.QPBT.MagicSquareRigidity.norm_applyOperatorToState_of_isometry hU]
end StateNorm
/-! ## Reflections
The *reflections* of `thm:ms-rigidity`, blueprint
`ch13_qpbt_test.tex:228-250`, are the self-adjoint involutions attached to
two-outcome projective measurements, which are exactly the binary observables
`MIPStarRE.QPBT.IsBinaryObservable` of
`MIPStarRE/QPBT/Combining/Linearity/Defs.lean`. The Magic Square rigidity
argument therefore uses that predicate and its algebra directly:
`IsBinaryObservable.one`, `IsBinaryObservable.mul`,
`IsBinaryObservable.isometry`, `isBinaryObservable_heteroKron_one` and
`isBinaryObservable_one_heteroKron`. -/
/-! ## Closeness in the state-dependent norm -/
/-- Two operators are *close within `δ` on `ψ`* when the state-dependent norm of
their difference is at most `δ`, that is, when the Euclidean distance between
`M ψ` and `N ψ` is at most `δ`.
This is **not** the relation `M ≈_δ N` of `thm:ms-rigidity`, blueprint
`ch13_qpbt_test.tex:249`, paper
`08_classical_and_quantum_low_degree_tests.tex:644-646`: there the relation
bounds the *squared* quantity `⟨ψ|(M - N)ᴴ (M - N)|ψ⟩` by `O(δ)`, and it is
formalized in this development by `opDistSq` (`def:povm-distance`, blueprint
`ch12_qpbt_games.tex:219-226`). The unsquared relation defined here is the one
under which the triangle inequality is additive, which is what the
solution-group computation of `Rigidity/Anticommutation.lean` iterates;
`NormCloseOn.opDistSq_le` converts a bound at scale `δ` here into the bound `δ ^ 2`
for `opDistSq`. -/
def NormCloseOn {ι : Type} [Fintype ι] [DecidableEq ι]
(ψ : EuclideanSpace ℂ ι) (δ : ℝ) (M N : _root_.LIDT.MIPStarRE.Quantum.Op ι) : Prop :=
‖_root_.MIPStarRE.QPBT.applyOperatorToState (M - N) ψ‖ ≤ δ
/-! ## Sign observables of a projective measurement -/
section SignObs
variable {α d : Type} [Fintype α] [DecidableEq α] [Fintype d] [DecidableEq d]
/-- The `±1`-valued observable attached to a measurement together with a binary
relabelling of its answers: an answer labelled `0` carries the eigenvalue `1`
and an answer labelled `1` carries the eigenvalue `-1`. For the Magic Square
these are the cell observables of `thm:ms-rigidity`, blueprint
`ch13_qpbt_test.tex:224-253`, paper
`references/qpbt-paper/08_classical_and_quantum_low_degree_tests.tex:612-652`. -/
noncomputable def signObs (M : _root_.LIDT.MIPStarRE.Quantum.Measurement α d) (f : α → ZMod 2) : _root_.LIDT.MIPStarRE.Quantum.Op d :=
∑ a, (_root_.MIPStarRE.QPBT.MagicSquareRigidity.bitSign (f a) : ℂ) • M.effect a
end SignObs
/-! ## Deviation of joint spectral combinations on the strategy state -/
section JointDefect
variable {G : _root_.MIPStarRE.QPBT.Game}
/-- Formalization-only: the joint effects of the two players form a projective
measurement on the composite space when both local measurements are
projective. -/
noncomputable def jointMeasurement (T : _root_.MIPStarRE.QPBT.Strategy G) (x : G.QuestionA)
(y : G.QuestionB) :
_root_.LIDT.MIPStarRE.Quantum.Measurement (G.AnswerA × G.AnswerB) (T.ιA × T.ιB) :=
_root_.LIDT.MIPStarRE.Quantum.Measurement.ofSumEqOne
(fun ab => _root_.MIPStarRE.QPBT.heteroKron ((T.A x).effect ab.1) ((T.B y).effect ab.2))
(fun ab => _root_.LIDT.MIPStarRE.Quantum.kronecker_nonneg ((T.A x).pos ab.1) ((T.B y).pos ab.2))
(by
rw [Fintype.sum_prod_type]
calc ∑ a : G.AnswerA, ∑ b : G.AnswerB,
_root_.MIPStarRE.QPBT.heteroKron ((T.A x).effect a) ((T.B y).effect b)
= ∑ a : G.AnswerA, _root_.MIPStarRE.QPBT.heteroKron ((T.A x).effect a)
(∑ b : G.AnswerB, (T.B y).effect b) := by
exact Finset.sum_congr rfl fun a _ => (_root_.MIPStarRE.QPBT.MagicSquareRigidity.heteroKron_sum_right _ _).symm
_ = _root_.MIPStarRE.QPBT.heteroKron (∑ a : G.AnswerA, (T.A x).effect a) (1 : _root_.LIDT.MIPStarRE.Quantum.Op T.ιB) := by
rw [(T.B y).sum_eq_one, _root_.MIPStarRE.QPBT.MagicSquareRigidity.heteroKron_sum_left]
_ = 1 := by rw [(T.A x).sum_eq_one, _root_.MIPStarRE.QPBT.heteroKron_one_one])
end JointDefect
end
end MIPStarRE.QPBT.MagicSquareRigidity
Source