Direct Polynomial Agreement Avg Le Mdq
ProvedMIPStarRE.QPBT.directPolynomialAgreement_avg_le_mdqTwo distinct direct polynomial representatives agree at a uniformly sampled direct point with probability at most m d / q.
All objects, parameters, typeclass assumptions, and hypotheses are quantified in the accompanying formal statement; each symbol denotes the object of the displayed Lean type with that name. The conclusion is the proposition
This result is an importable reduction step in the formal dependency graph for the quantum Pauli basis test completeness or soundness argument.
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.
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_AxisParallelLine
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_DiagonalLine
import Definitions.Def_LIDT_MIPStarRE_LDT_Basic_Distribution
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_Preliminaries_Polynomials
import Definitions.Def_LIDT_MIPStarRE_LDT_Tactic_AvgCongr
import Definitions.Def_LIDT_MIPStarRE_LDT_Test_StrategyCore
import Definitions.Def_QPBT_QPBT_Algebra_FieldBasis
import Definitions.Def_QPBT_QPBT_Combining_DirectLowDegree_Geometry
import Definitions.Def_QPBT_QPBT_Combining_DirectLowDegree_Transport_Consistency_Measurements
import Definitions.Def_QPBT_QPBT_Combining_DirectLowDegree_Transport_Correspondence
import Definitions.Def_QPBT_QPBT_Test_LowDegreeGameMeasurements
import Mathlib
import Mathlib.Analysis.MeanInequalitiesPow
namespace LIDT.MIPStarRE.LDT
end LIDT.MIPStarRE.LDT
namespace LIDT.MIPStarRE.LDT.Preliminaries
end LIDT.MIPStarRE.LDT.Preliminaries
namespace LIDT.MIPStarRE.Quantum
end LIDT.MIPStarRE.Quantum
namespace MIPStarRE.QPBT
end MIPStarRE.QPBT
/-!
# Simultaneous polynomial measurements for the direct low-degree game
This module applies the quantum soundness theorem of the low individual
degree test to each coordinate of a projective strategy for the directly
indexed low-degree game, transports the Schwartz--Zippel collision estimate
to the direct polynomial representatives, and, for simultaneity parameter
`1`, obtains the polynomial-tuple conclusion of `lem:ld-soundness` from the
coordinate conclusions.
## The simultaneity obstruction
For simultaneity parameter `k ≥ 2` the coordinate conclusions alone do not
determine simultaneous polynomial measurements, so no palindromic combination
of the coordinate measurements can be consistent with a joint point
measurement in general. Let `q = 2^s`, let the degree be `1`, and let Bob
measure the univariate polynomials `g = g₀ + g₁ x` once in the standard basis
`|g⟩` and once in the Fourier basis attached to the symmetric pairing
`⟨h, g⟩ = h₀ g₁ + h₁ g₀`. In characteristic two every evaluation functional
`h ↦ h(u)` is represented by an isotropic vector of this pairing, so at every
point `u` the value-level coarse-grainings `Q^{1,u}_{[a]}` (Fourier basis) and
`Q^{2,u}_{[b]}` (standard basis) commute, and on the maximally entangled state
Alice may answer point questions with the joint projective measurement
`P^u_{a,b} = (Q^{1,u}_{[a]} Q^{2,u}_{[b]})ᵀ`. All coordinate consistencies of
`lem:ld-soundness` then hold with defect `0` and distinct polynomials collide
at a uniform point with probability `1 / q`, but the sandwiched POVM has
`∑_{g₂(u) = b} |g₂⟩⟨g₂| Q^{1,u}_{[a]} |g₂⟩⟨g₂| = q⁻¹ Q^{2,u}_{[b]}`, whose
consistency defect against `P^u` is `1 - 1/q`; more strongly, every
polynomial-pair POVM on Bob's side has defect at least `1 - 2/q` against
`P^u`. No bound of the form `C k √(δ + m d / q)` holds for a universal `C`.
The source obtains the general case not coordinatewise but by the combining
reduction of Theorem 4.43 in the NEEXP paper, which applies the `k = 1`
theorem once, in dimension `m + k`, to a combined strategy; that reduction is
not formalized here. The two-measurement pasting estimate (Fact 4.35 there)
does not apply either, since it evaluates the two coordinates at
independently sampled points while the low-degree game evaluates all
coordinates at one point. The counterexample and the resulting formal
status are recorded in `docs/paper-gaps/qpbt_ld-simultaneous-sandwich.tex`.
## References
* `references/qpbt-paper/08_classical_and_quantum_low_degree_tests.tex:413-458`
* `references/qpbt-paper/06_nonlocal_games_and_mipstar.tex:465-501`
* `references/neexp-paper/05_quantum_preliminaries.tex:912-1060` and `:1409-1503`
* `references/ldt-paper/test_definition.tex:180-202`
* `lem:ld-soundness` in `blueprint/src/chapter/ch13_qpbt_test.tex`
* `docs/paper-gaps/qpbt_ld-dimension-divisibility.tex`
* `docs/paper-gaps/qpbt_ld-simultaneous-sandwich.tex`
-/
open scoped BigOperators Matrix MatrixOrder ComplexOrder
open MIPStarRE.QPBT
open LIDT.MIPStarRE.LDT
open LIDT.MIPStarRE.LDT.Preliminaries
open LIDT.MIPStarRE.Quantum
/-! ## Direct polynomial collision estimate -/theorem MIPStarRE.QPBT.directPolynomialAgreement_avg_le_mdq (D : _root_.MIPStarRE.QPBT.DirectLdParams)
(g g' : _root_.MIPStarRE.QPBT.PolyIndex D.m (_root_.MIPStarRE.QPBT.DirectScalarQ D) D.d) (hneq : g ≠ g') :
_root_.LIDT.MIPStarRE.LDT.avgOver (_root_.LIDT.MIPStarRE.LDT.uniformDistribution (Fin D.m → _root_.MIPStarRE.QPBT.DirectScalarQ D))
(fun u => if MvPolynomial.eval u g.1 = MvPolynomial.eval u g'.1 then
(1 : _root_.LIDT.MIPStarRE.LDT.Error) else 0) ≤
(D.m * D.d : _root_.LIDT.MIPStarRE.LDT.Error) / D.q := by sorry