Lemma 7: Space-Time Decoupling #
Statement #
Any spacetime logical fault (Def_11) is equivalent, up to multiplication by spacetime stabilizers (Def_11), to the product of a space-only logical fault and a time-only logical fault.
Formally: Let F be a spacetime logical fault in the fault-tolerant gauging measurement procedure (Def_10). Then there exist:
- A space logical fault F_S: a spacetime fault consisting solely of Pauli errors at time t_i, such that F_S is a logical operator of the deformed code (Def_4).
- A time logical fault F_T: a spacetime fault consisting solely of time-faults (measurement/initialization errors), with no Pauli errors.
- A spacetime stabilizer S (Def_11).
such that F = F_S · F_T · S (composition via symmetric difference of fault sets).
Main Results #
FullOutcomePreserving: The full outcome predicate (sign + code state)IsSpaceLogicalFault: A space-only fault at t_i that is a logical of the deformed codeIsTimeLogicalFault: A pure-time fault that flips the gauging signspaceTime_decomposition: The main decomposition theorem F = F_S · F_T · Sspace_time_independent_effects: F_S and F_T affect different aspects of the outcome
Proof Outline #
- Move all space-faults to time t_i using time-propagating stabilizers (Lem 5)
- Convert boundary init/readout faults to space-faults at t_i using Lem 5 generators
- Separate the remaining time-faults into A_v strings (logical) or stabilizer generators
- Verify independence: space affects code state, time affects measurement sign
Part I: Full Outcome Predicate #
The paper's Def_11 says a spacetime logical fault "changes the outcome" in EITHER of two ways:
- Flipping the inferred measurement sign σ (time-type logical), OR
- Applying a nontrivial logical operator to the post-measurement code state (space-type logical).
A fault is outcome-preserving if NEITHER occurs: the sign is preserved AND the net Pauli error at t_i is in the deformed stabilizer group (not a nontrivial logical).
The deformed stabilizer code at the gauging time, constructed from proc's data.
Equations
- SpaceTimeDecoupling.theDeformedCode proc = DeformedCodeChecks.deformedStabilizerCode G cycles checks proc.deformedData ⋯ ⋯
Instances For
The full outcome-preserving predicate for the fault-tolerant gauging procedure. A fault preserves the outcome if BOTH: (1) The gauging sign σ is preserved (no time-logical effect), AND (2) The net Pauli error at t_i is in the deformed stabilizer group (no space-logical effect).
This captures the paper's Def_11: a spacetime stabilizer leaves both the classical measurement record and the quantum code state unchanged.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A spacetime logical fault under the full outcome predicate: syndrome-free AND changes the outcome (either flips sign or applies nontrivial logical).
Equations
Instances For
A spacetime stabilizer under the full outcome predicate: syndrome-free AND preserves both the sign and code state.
Equations
Instances For
A logical fault changes the outcome: ¬(sign preserved ∧ Pauli in stabilizer group). Equivalently: either the sign is flipped, or the Pauli error is a nontrivial logical.
Part II: Space-Only and Time-Only Fault Definitions #
A space-only logical fault: all space-faults are concentrated at t_i, no time-faults, and the composite Pauli error at t_i is a nontrivial logical of the deformed code.
Equations
- One or more equations did not get rendered due to their size.
Instances For
A time-only logical fault: no space-faults, syndrome-free, and flips the gauging sign σ.
Equations
- SpaceTimeDecoupling.IsTimeLogicalFault proc F = (F.isPureTime ∧ SpacetimeLogicalFault.IsSyndromeFreeGauging proc proc.detectorOfIndex F ∧ SpacetimeLogicalFault.FlipsGaugingSign proc F)
Instances For
Part III: Z₂ Algebra of Symmetric Differences #
The key algebraic fact: sums of ZMod 2 indicators over symmetric differences decompose additively. This underlies all composition properties of the gauging procedure.
Part III-A: Composition Properties — gaussSignFlip #
The gaussSignFlip is a double sum of ZMod 2 indicators over timeFaults. Since composition takes the symmetric difference of timeFaults, the Z₂ additivity of indicators gives additivity of gaussSignFlip.
The gaussSignFlip is Z₂-additive under composition: sign(F₁ · F₂) = sign(F₁) + sign(F₂) in ZMod 2. This follows from the fact that timeFaults of the composition is the symmetric difference, and the indicator function is additive over Z₂.
Part III-B: Composition Properties — pauliErrorAt #
The Pauli error at time t is defined via sums over spaceFaultsAt t. Since composition takes the symmetric difference of spaceFaults, and filter distributes over symmDiff, the Z₂ sum additivity gives multiplicativity of PauliOp.
The net Pauli error at t is multiplicative under composition: pauliErrorAt(F₁ · F₂, t) = pauliErrorAt(F₁, t) * pauliErrorAt(F₂, t). This follows because spaceFaults of the composition is the symmetric difference, filter distributes over symmDiff, and Z₂ sums are additive = Pauli product.
Part III-C: Composition Properties — Syndrome-Freeness #
Detector violation depends on flipParity, which is a ZMod 2 sum of indicators over timeFaults. By the same Z₂ additivity, composing two syndrome-free faults preserves syndrome-freeness.
Composing two syndrome-free faults preserves syndrome-freeness. Both faults have empty syndrome, so the composed fault (symmetric difference of timeFaults) also has empty syndrome by Z₂ additivity of flipParity.
Composing with a full stabilizer preserves syndrome-freeness.
Part III-D: Composition Properties — Full Stabilizer Closure #
Combining the above: the composition of two full stabilizers is a full stabilizer. Syndrome-freeness, sign preservation, and stabilizer group membership all transfer.
The composition of two full gauging stabilizers is a full gauging stabilizer. Syndrome-freeness and outcome preservation are both additive over Z₂.
Part IV: Time-Propagating Stabilizers Move Space-Faults #
The time-propagating generators from Lem_5 allow moving a Pauli error from time t to time t+1 (or t-1). By composing these, we can move any space-fault to the gauging time t_i = phase2Start.
The Empty Fault is a Full Gauging Stabilizer #
The empty fault (no space-faults, no time-faults) trivially satisfies all stabilizer conditions: empty syndrome, zero sign flip, and the identity Pauli is in the stabilizer group.
The empty fault is a full gauging stabilizer.
Boundary faults can be absorbed: if space-faults are already concentrated at t_i,
the empty fault serves as a trivial stabilizer preserving this property.
Steps 1 and 4 of the paper are subsumed by space_fault_cleaning (Lem 5),
which handles both time-propagation and boundary absorption in one step.
Space-Fault Cleaning and Centralizer Membership #
Using the space_fault_cleaning and syndromeFree_pureSpace_inCentralizer axioms from Lem 5, we prove the two key constructive lemmas needed for the main decomposition:
- Any syndrome-free fault can be cleaned to have space-faults only at t_i
- A pure-space fault concentrated at t_i has its Pauli error in the centralizer
Space-fault cleaning (Steps 1+4): Any syndrome-free spacetime fault can be composed with a full gauging stabilizer to concentrate all space-faults at t_i. This uses the time-propagating and boundary generators from Lem 5.
Centralizer membership (Step 3): A syndrome-free pure-space fault concentrated at t_i has its Pauli error in the centralizer of the deformed code. This encodes the quantum-mechanical fact that the deformed code checks are the active checks at t_i, and the cleaning process preserves commutation.
Part V: Time-Fault Classification #
After cleaning, the time-faults form a pure-time fault. By Lem 6, syndrome-free pure-time faults either flip σ (logical time-fault = A_v string) or decompose into stabilizer generators (trivial).
The cleaned time-faults of a syndrome-free fault either flip the sign or are stabilizers.
This is the content of Lem 6's pureTime_syndromeFree_logical_or_stabilizer_generators.
Part VI: Independence of Space and Time Effects #
The space fault F_S and time fault F_T affect the procedure outcome independently:
- F_S acts on the code Hilbert space (applies a Pauli logical operator)
- F_T acts on the classical measurement record (flips σ)
The gauging sign depends only on time-faults (measurement errors on Gauss checks). Space-faults do not affect the gauging sign directly.
A pure-space fault preserves the gauging sign.
The Pauli error at any time depends only on space-faults. Two faults with the same spaceFaults have the same pauliErrorAt.
A pure-time fault does not change the Pauli error at any time.
Independence theorem: The space and time components of a fault affect different aspects of the procedure outcome.
- The gauging sign σ depends only on time-faults (measurement errors).
- The Pauli error on the code state depends only on space-faults. Therefore F_S and F_T have independent effects.
The sign flip of F_S · F_T equals the sign flip of F_T when F_S is pure-space.
Part VII: Main Decomposition Theorem #
The central result: any spacetime logical fault decomposes as F = F_S · F_T · S.
Lemma 7 (Space-Time Decoupling): Main Theorem.
Any spacetime logical fault F in the fault-tolerant gauging measurement procedure is equivalent, up to multiplication by spacetime stabilizers, to the product of a space-only fault F_S and a time-only fault F_T.
Specifically: there exist F_S, F_T, and S such that:
- F = F_S · F_T · S (composition via symmetric difference)
- F_S is pure-space (no time-faults) and concentrated at time t_i
- F_T is pure-time (no space-faults) and syndrome-free
- S is a full gauging stabilizer (syndrome-free, outcome-preserving)
- If F_T is nontrivial, it flips the gauging sign σ (time-logical)
- If F_S is nontrivial, its Pauli error at t_i is a nontrivial deformed-code logical
The proof proceeds by:
- Move all space-faults to t_i using time-propagating stabilizers (Lem 5)
- Convert boundary init/readout faults to space-faults at t_i (Lem 5 generators)
- Separate the cleaned fault into space and time parts
- Classify the time part via Lem 6's dichotomy
Part VIII: The A_v String as Canonical Time Logical Fault #
The A_v string is the canonical time-only logical fault. It has weight d, is syndrome-free, and flips the gauging sign when d is odd.
Time-only logical faults have weight at least d (when d is odd).
Sign flip of the A_v string.
Part IX: Sign-Flip Analysis #
A pure-space fault has zero sign flip.
Part X: Structural Decomposition Properties #
Any spacetime fault decomposes into its space and time components.
The weight of a fault is the sum of its space and time weights.
The compose of pure-space and pure-time faults preserves the space component.
The compose of pure-space and pure-time faults preserves the time component.
When F_S is pure-space and F_T is pure-time, their composition has weight equal to the sum of their weights (since their fault sets are disjoint).
Part XI: Decoupling Summary #
Full summary combining decomposition with independence of effects.
If F is a logical fault that flips the sign, then the time component contains at least one full A_v measurement string.
Main summary: Space-time decoupling for the fault-tolerant gauging procedure.
For any spacetime logical fault F of the fault-tolerant gauging measurement procedure:
- F decomposes as F = F_S · F_T · S where:
- F_S is pure-space (Pauli errors only), concentrated at t_i
- F_T is pure-time (measurement errors only), syndrome-free
- S is a full gauging stabilizer (preserves sign AND code state)
The gauging sign σ is determined entirely by F_T: gaussSignFlip(F) = gaussSignFlip(F_T)
The Pauli error on the code state is determined entirely by F_S: pauliErrorAt(F, t) = pauliErrorAt(F_S, t) for all t
If F flips the sign, then F_T is a time-only logical fault (A_v measurement string with weight d, characterized in Lem 6)
If F applies a nontrivial space logical, then F_S is a space logical fault (deformed code logical operator at time t_i)
Part XII: Weight Bounds from Decoupling #
Weight bound from decoupling: For any pair of pure-space and pure-time faults, their composition has weight equal to the sum of weights.