§C.1, — exchange gives on alone and on alone
ProvedBaconThesis.ExchangeUniv.n3_independentdecoherence-free-subspaceslie-algebrasp2o-gran-per-chapterp2o-plan-bookp2o-v1quantum-information
Let be the real Lie algebra generated by on three qubits, and let , be the total-spin- and subspaces.
- contains an operator that acts as on and as on its orthogonal complement; so contains acting on alone.
- Every traceless anti-Hermitian that commutes with and annihilates the orthogonal complement of lies in .
An operator commuting with the collective spins and supported on has the form , with ; so clause 2 says that contains acting on and as zero elsewhere. This is the base case of the induction of Appendix C.
Formalization Note is exchangeLie 3; Hamiltonians are represented by .
Differences from the source:
- reading: "we can perform U(1) on DFS3(3/2) alone" → L contains an element acting as i·I on the total-spin-3/2 subspace and 0 on its complement; reason: that element spans u(1) on that block; the unitaries are its exponentials
- reading: "SU(2) on DFS3(1/2) alone" (gates, by an Euler angle construction) → every element of su(2) ⊗ I₂ on the total-spin-1/2 block, extended by zero, lies in L; reason: the connected group of such gates has exactly this Lie algebra, and the page obtains the gates from the Lie algebra elements σ_z, σ_x of (C.4)
Preamble
import Mathlib import Definitions.Def_BaconThesis_ExchangeUniv_collS import Definitions.Def_BaconThesis_ExchangeUniv_exchangeLie import Definitions.Def_BaconThesis_ExchangeUniv_dfsSpace import Definitions.Def_BaconThesis_ExchangeUniv_suOn open scoped Matrix
Formal statement
attribute [local instance 100] LieRing.ofAssociativeRing
namespace BaconThesis.ExchangeUniv
/-- **§C.1, `n = 3`: `U(1)` on `DFS₃(3/2)` alone and `SU(2)` on `DFS₃(1/2)` alone.**
Bacon, *Decoherence, Control, and Symmetry in Quantum Computers*, arXiv:quant-ph/0305025v1, §C.1, after (C.4), p.246 (PDF p.263): "it is possible to act independently on DFS₃(3/2)
and DFS₃(1/2). In other words, we can perform U(1) on DFS₃(3/2) alone, and SU(2) on DFS₃(1/2)
alone."
1. The BaconThesis.Collective.exchange Lie algebra `L` of three qubits contains an element acting as `i·I` on the
total-spin-`3/2` subspace and as `0` on its orthogonal complement (so `L` contains
`u(1) = iℝ` on `DFS₃(3/2)` alone).
2. Every traceless anti-Hermitian `X` commuting with `S_x, S_y, S_z`, acting only on the
total-spin-`1/2` subspace (zero on its orthogonal complement), lies in `L`: this is
`su(2) ⊗ I₂` on `DFS₃(1/2)` alone.
**Formalization Note.** `L = exchangeLie 3`, the real Lie span of the `i·E_ij`; an operator
commuting with every `S_α` and supported on the spin-`1/2` block is `Y ⊗ I₂` in the block, and
`tr X = 2 tr Y`, so clause 2 is exactly `su(2) ⊗ I₂` extended by zero. -/
theorem n3_independent :
(∃ X ∈ exchangeLie 3,
(∀ v ∈ dfsSpace 3 (3 / 2), X *ᵥ v = Complex.I • v) ∧
(∀ v : (Fin 3 → Fin 2) → ℂ, (∀ w ∈ dfsSpace 3 (3 / 2), star w ⬝ᵥ v = 0) → X *ᵥ v = 0)) ∧
(∀ X : Matrix (Fin 3 → Fin 2) (Fin 3 → Fin 2) ℂ,
X ∈ suOn (dfsSpace 3 (1 / 2)) → (∀ α : Fin 3, Commute X (collS 3 α)) →
X ∈ exchangeLie 3) := by sorry
end BaconThesis.ExchangeUniv
Source
Bacon, Decoherence, Control, and Symmetry in Quantum Computers, arXiv:quant-ph/0305025v1, §C.1, after (C.4), p.246 (PDF p.263)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.