Here we study the unitary groups that can be constructed using elements from the qubit Clifford Hierarchy. We first provide a necessary and sufficient canonical form that semi-Clifford and generalized semi-Clifford elements must satisfy to be in the Clifford Hierarchy. Then we classify the groups that can be formed from such elements. Up to Clifford conjugation, we classify all such groups that can be constructed using generalized semi-Clifford elements in the Clifford Hierarchy. We discuss a possible minor exception to this classification in the appendix. This may not be a full classification of all groups in the qubit Clifford Hierarchy as it is not currently known if all elements in the Clifford Hierarchy must be generalized semi-Clifford. In addition to the diagonal gate groups found by Cui et al., we show that many non-isomorphic (to the diagonal gate groups) generalized symmetric groups are also contained in the Clifford Hierarchy. Finally, as an application of this classification, we examine restrictions on transversal gates given by the structure of the groups enumerated herein which may be of independent interest.
A gate in U(2^n) is semi-Clifford iff it can be written as C_1 \prod_j\exp \left( i\theta_j Z_j \right) C_2 where C_1,C_2 are n-qubit Clifford gates, \theta \in [0,2\pi), and Z_j are non-identity Z Pauli strings.
A gate in U(2^n) is generalized semi-Clifford iff it can be expressed as
where C_1,C_2 are n-qubit Clifford gates, \theta \in [0,2\pi), Z_j are Z Pauli strings, and P is a permutation matrix in U(2^n).
All elements in the qubit Clifford Hierarchy are generalized semi-Clifford.
A set of semi-Clifford gates in \mathcal{CH} generates a group of semi-Clifford gates in \mathcal{CH} iff each element, l, in the set, written in canonical form as C_l \prod_{j_l}\exp \left( i\frac{\alpha_{j_l} \pi}{2^{k_{j_l}}}S_{j_l} \right), has all S_{j_l} in a common stabilizer group, S, and has all C_l in the normalizer of S in the n-qubit Clifford group.
For every stabilizer group on n qubits, there exists an n-qubit Clifford circuit which by conjugation takes all elements of S to Pauli Z strings (a diagonal stabilizer group). Furthermore, if the stabilizer group is on n qubits, but has 2^m elements for some m<n, then a Clifford circuit exists which by conjugation takes the elements of S to Pauli Z strings on m qubits and to Identity on the remaining n-m qubits.
Every group, G, of semi-Clifford gates on n qubits in \mathcal{CH} must be Clifford-isomorphic to a subgroup of quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a () command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of Up to certain permutations described in Appendix ?.
A permutation group on n qubits, \tilde{\Pi}^n, is maximal in \mathcal{CH} if all elements P \in\tilde{\Pi}^n are also in \mathcal{CH} and the addition of any permutation P' \in \Pi^n and P' \notin \tilde{\Pi}^n results in a group which is no longer contained in \mathcal{CH}. We refer to these groups as maximal permutation groups for short.
Every group, G, of generalized semi-Clifford gates on n qubits in \mathcal{CH} must be Clifford-isomorphic to a subgroup of quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a quantumarticleYou must not put a () command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of command inside the argument of Up to certain permutations described in Appendix ?.
Let U\in G_{l}^n, then UpU^\dagger \in G_{l}^n for any n-qubit Pauli string, p.
Let \pi d \in G_l^n and \pi d \in \mathcal{CH}_1 (the Pauli group). Then, \pi = p_X for some Pauli X string and d = p_Z for some Pauli Z string.
Let \pi_k be a permutation matrix in \mathcal{CH}_k, then \pi_k p_X \pi_k^\dagger is a permutation matrix in \mathcal{CH}_{k-1} for all Pauli X strings, p_X. And, \pi_k p_Z \pi_k^\dagger is a gate in Diag_1 in \mathcal{CH}_{k-1} for all Pauli Z strings, p_Z.
U\in \mathcal{CH}_k, iff C_1 U C_2 \in \mathcal{CH}_k, where C_1, C_2 are arbitrary Clifford gates.
Any d\in\mathrm{Diag}^n_l is in the Clifford Hierarchy at some level. This is the case, since it can be expressed as d=\prod_j\exp \left( i\frac{\alpha_j \pi}{2^{k_j}}Z_j \right) where Z_j are Pauli Z strings and \alpha_j are integers. When terms with the same Pauli strings are combined and the fractions \frac{\alpha_j}{2^{k_j}} are reduced, the level of d is the maximum k_j.
Conjugation of an element d\in \mathrm{Diag}^n_l by an element \pi\in\Pi^n yields a diagonal gate in \mathcal{CH} at some (possibly higher) level. \pi d \pi^{-1} must be in \mathrm{Diag}^n_l since \pi is in the normalizer of \mathrm{Diag}^n_l. Furthermore, it must be in \mathcal{CH} at some level by Lemma 1.5.
where \theta \in [0,2\pi) and P_1,P_2 \in \mathcal{P}, the n-qubit Pauli group.
Note that e^{i\theta P} = I^{\otimes n} \cos{\theta} + iP\sin{\theta} and verification of the above equation is straightforward.
Conjugation of an n-qubit Pauli string, p, by a diagonal gate (on n qubits) d_k in \mathcal{CH}_k yields d_kpd_k^{-1} = pd_{k-1} \in \mathcal{CH}_{k-1}. To see this, note that d=\prod_j\exp \left( i\frac{\alpha_j \pi}{2^{k_j}}Z_j \right) with k_j \le k for all j. Also, note that \prod_j\exp \left( i\frac{\alpha_j \pi}{2^{k_j}}Z_j \right) p = p\prod_j\exp \left( \pm i\frac{\alpha_j \pi}{2^{k_j}}Z_j \right) via Lemma 1.7 where +(-) indicates that p commutes (anticommutes) with Z_j. However a commuting term will be cancelled by its corresponding term in d^{-1} and an anticommuting term will combine with its corresponding term in d^{-1} picking up a factor of 2. Since all k_j are now reduced by at least one, we conclude that pd_{k-1} \in \mathcal{CH}_{k-1}.
If a permutation gate, \pi\in\mathcal{CH}, and a diagonal gate, d \in \mathcal{CH}, then \pi d \in \mathcal{CH}.
Base case k=2: For k\le 2 all permutations are Clifford and by Lemma 1.4, any Clifford times a (diagonal) gate in \mathcal{CH}_k is in \mathcal{CH}_k. Therefore, all \pi D \in G^n[2] are in \mathcal{CH} since all D \in \mathcal{CH} (at some level).
Assume, \pi d \in G^n[k-1] \implies \pi d \in \mathcal{CH}.
For elements \pi_k D \in G^n[k] we have that \pi_k p \pi_k^\dagger \in \mathcal{CH}_{k-1} and by Lemma 1.3 we know this implies that \pi_k p_X \pi_k^\dagger is a permutation in \mathcal{CH}_{k-1} for all p_X and \pi_k p_Z \pi_k^\dagger is a diagonal gate in \mathcal{CH}_{k-1} for all p_Z.
To prove that \pi_k D \in \mathcal{CH}, we must show that \pi_k D p D^\dagger \pi_k^\dagger is in \mathcal{CH} for all p. We can write this as: \pi_k D p D^\dagger \pi_k^\dagger = \pi_k p D' \pi_k^\dagger (where D' is a diagonal gate in \mathcal{CH}) by Lemma 1.8. This equals, \pi_k p_X \pi_k^\dagger (\pi_k p_Z \pi_k^\dagger \pi_k D' \pi_k^\dagger) where the product of all gates in the parentheses is a diagonal gate in \mathcal{CH} (for all p_Z) by Lemma 1.6 and by closure of the group Diag_l^n (for some, large enough, l). We will denote this product as D'' in what follows. By the discussion in the previous paragraph, we know that \pi_k p_X \pi_k^\dagger is a permutation in \mathcal{CH}_{k-1} for all p_X. Finally, since \pi_k p_X \pi_k^\dagger D'' = \pi_{k-1} D'' \in G^n[k-1] (for all p_X) which by our inductive hypothesis is in \mathcal{CH}, we conclude that any element \pi_k D \in G^n[k] is in \mathcal{CH}.
When checking if a permutation, \pi, is in \mathcal{CH}, it suffices only to check if \pi p_X \pi^{-1} \in \mathcal{CH} for all p_X.
If a permutation gate, \pi\in\mathcal{CH}, and a diagonal gate, d \notin \mathcal{CH}, then \pi d \notin \mathcal{CH}.
Base case k=2: Here, \pi is Clifford and d\notin \mathcal{CH}. Assume that g=\pi d \in \mathcal{CH}. Then, d = \pi^\dagger g, but \pi^\dagger is a Clifford gate since the Clifford gates form a group, and \pi^\dagger g \in \mathcal{CH} since multiplication by Clifford gate does not change the level in \mathcal{CH}. We then have that d\in \mathcal{CH}, a contradiction. We conclude that all elements in \tilde{G}^n[2] are not in \mathcal{CH}.
Assume, \pi d \in \tilde{G}^n[k-1] \implies \pi d \notin \mathcal{CH}.
Let \pi d be an element in \tilde{G}^n[k]. We want to show that there exists some Pauli string p such that \pi d p d^\dagger \pi^\dagger \notin \mathcal{CH}. Since d\notin \mathcal{CH}, there must exist some p_X (since all p_Z commute with d) such that dp_X d^\dagger \notin \mathcal{CH} and this equals p_X d' for some diagonal matrix d' (this is a consequence of Lemma. 1.7). In what follows, p_X will denote this specific Pauli X string which we will use to show that \pi d \notin \mathcal{CH}. We can write \pi d p_X d^\dagger \pi^\dagger = \pi p_X \pi^\dagger \pi d' \pi^\dagger. But, \pi \in \mathcal{CH}_k which implies that \pi p_X \pi^\dagger is a permutation in \mathcal{CH}_{k-1}. We conclude that \pi p_X \pi^\dagger \pi d' \pi^\dagger \in \tilde{G}^n[k-1] and by our inductive hypothesis, we have that \pi p_X \pi^\dagger \pi d' \pi^\dagger \notin \mathcal{CH} and therefore \pi d \notin \mathcal{CH}.
If \pi \notin \mathcal{CH} then \pi d \notin \mathcal{CH}.
We prove this theorem by contradiction.
Let \pi_1 \notin \mathcal{CH}. Assume that \pi_1 d_1 \in \mathcal{CH}_k for an arbitrary diagonal gate, d_1.
Since \pi_1 \notin \mathcal{CH} we have that there exists a p^{1,2}_X such that \pi_1 p^{1,2}_X \pi_1^\dagger = \pi_2 \notin \mathcal{CH} where \pi_2 is a permutation not necessarily distinct from \pi_1. Such a p^{1,2}_X and \pi_2 must exist or \pi_1 would be in \mathcal{CH}. And since \pi_2 \notin \mathcal{CH} we also have that \pi_2 p^{2,3}_X \pi_2^\dagger = \pi_3 \notin \mathcal{CH}. And more generally, \pi_t p^{t,t+1}_X \pi_t^\dagger = \pi_{t+1} \notin \mathcal{CH}. These permutations are not necessarily distinct and the progression may ‘loop’, but we can continue the progression (even if a loop occurs) to arbitrary length t, and each \pi_t will not be in \mathcal{CH}.
Now, if \pi_1 d_1 \in \mathcal{CH}_k we have that \pi_1 d_1 p d_1^\dagger \pi_1^\dagger \in \mathcal{CH}_{k-1} for all Pauli strings p. But p^{1,2}_X is one such Pauli string and therefore
where d_1' is given by Lemma 1.8 and d_2 = \pi_1 d_1' \pi_1^\dagger. And by assumption \pi_2 d_2 \in \mathcal{CH}_{k-1}. We can proceed in this manner choosing p_X^{t,t+1} since it must be true for all Pauli strings, p. Whatever level k we started with, we will eventually get to level k=1 and there we have that \pi_t d_t \in \mathcal{CH}_1. Since this must be a Pauli group element, P, which can be expressed (up to a phase) as a product of p_X p_Z, where p_Z is a diagonal Pauli gate, and p_X is a permutation (Pauli) gate, we have that \pi_t d_t = p_X p_Z and via the splitting property \pi_t = p_X and d_t = p_Z. But \pi_t \notin \mathcal{CH} (at any level) and p_X \in \mathcal{CH}_1. We have a contradiction and our assumption that \pi_1 d_1 \in \mathcal{CH}_k must not be true. Therefore, \pi_1 d_1 \notin \mathcal{CH}_k. But since this proof works for arbitrary k, we conclude that \pi_1 d_1 \notin \mathcal{CH}.
If \pi d \in \mathcal{CH}_k, then \pi \in \mathcal{CH}_k.
If \pi d \in \mathcal{CH}_2, then \pi \in \mathcal{CH}_2 and d \in \mathcal{CH}_2.
If a permutation can be written as a circuit, U, such that the CT mismatch is zero between any two time slices, then U is in \mathcal{CH}.
Note that CT mismatch of zero is equivalent to requiring that all time slices commute. Then, since any time slice can be diagonalized by conjugation with local Clifford gates, simply conjugate each target by H. And, since the CT mismatch is zero, all time slices can be diagonalized by the same local Clifford gates; U is therefore Clifford-isomorphic to a diagonal gate element in \mathcal{CH} which shows that U must be in \mathcal{CH}.
A Pauli X string can be ‘passed through’ a permutation gate, C^n(X), without increasing the CT mismatch in the circuit.
We will prove the result for a weight-one Pauli X string, and note that by repeating the result (with other weight-one Pauli X strings), it can be applied to an arbitrary X string. Using the following circuit identity:
we see that the Pauli X string can be passed through any C^n(X) gate without increasing the CT mismatch. The additional C^{n-1}(X) gate has targets and controls at the same locations as the C^n(X) gate and will have the same or lower CT mismatch with all other gates in the circuit. Note that this circuit identity, while derived independently, is similar to the ‘Pauli Sandwich Trick’ of O’Connor and Yoder[OConnor:2021a].
Since multiplication of U on the left and right by Clifford permutations (Xs and CNOTs) does not change the level in \mathcal{CH} we can extend Theorem 4.1 to allow arbitrary Clifford permutations on the left and right of U. Also, since Pauli X strings can be ‘passed through’ a permutation circuit without increasing the CT mismatch, we can allow arbitrary X strings between time slices.
Any permutation circuit with CT mismatch zero is in \mathcal{CH} at the level of the highest-level gate in the circuit.
It is well known that C^n(X) is in \mathcal{CH} at level n+1 [Zeng2008]. We only need to check Pauli X strings by Lemma 1.9.1. By Lemma 4.2 we see that we can pass Pauli X strings through permutations and the resulting permutations must commute with all other permutations. This allows us to treat each time slice separately, as well as each gate in a time slice if there is more than one. Then, the permutation circuit is in \mathcal{CH} at the level of the highest-level individual gate, as claimed.
A permutation is in \mathcal{CH} at the third level iff it can be written as a circuit of commuting Toffoli (C^2(X)) gates (possibly preceded and followed by Clifford permutations).
If a permutation \pi is in \mathcal{CH}_k, then \pi^\dagger is in \mathcal{CH}_k.
For any d \in \mathcal{D}_k^n, d^2 \in \mathcal{D}_{k-1}^n.
For any diagonal gates d'\notin \mathcal{CH} and d\in\mathcal{CH}, d'd \notin \mathcal{CH}.
For any d\notin \mathcal{CH}, we have that \pi d \pi^\dagger \notin \mathcal{CH} for any permutation \pi.
For any d\notin \mathcal{CH}, we have that d^2 \notin \mathcal{CH}.
For any d \in Diag_k^n, d^2 \in Diag_{k-1}^n.
The text on this page is from arXiv:2212.05398v2 by Jonas T. Anderson, under the licence CC BY 4.0. The corpus made these changes: it read the items out of the LaTeX source, it expanded the macros of the paper, it put each proof after its statement, and it converted the text to HTML. The formulas are the TeX of the paper, and KaTeX renders them. A formula or a reference can look different from the PDF. The page is published under CC BY 4.0.
The record of each item is in analysis/objects/2212/2212.05398.jsonl.