Magic informationally complete POVMs with permutations


Eigenstates of permutation gates are either stabilizer states (for gates in the Pauli group) or magic states, thus allowing universal quantum computation [M. Planat and Rukhsan-Ul-Haq, Preprint 1701.06443]. We show in this paper that a subset of such magic states, when acting on the generalized Pauli group, define (asymmetric) informationally complete POVMs. Such IC-POVMs, investigated in dimensions $2$ to $12$, exhibit simple finite geometries in their projector products and, for dimensions $4$ and $8$ and $9$, relate to two-qubit, three-qubit and two-qutrit contextuality.
Submitted 10 Apr 2017 to Quantum Physics [quant-ph]
Published 11 Apr 2017
Updated 22 Aug 2017
Author comments: 11 pages, 6 figures, 1 table, a few complements and corrections
Msc class: 81P50, 81P68, 81P13, 81P45, 20B05
Journal ref: R. 4:170387(2017)


Blake Stacey Apr 12 2017 00:34 UTC

Clarification request: Are all the IC-POVMs in this paper minimal? That is, does the number of elements in each POVM equal the square of the dimension? If so, I am confused about the quoted value of the inner product between projectors for the equiangular IC-POVM in dimension 5.

Planat Apr 12 2017 13:58 UTC

Yes, the IC-POVMs under consideration are minimal. The IC-POVM in dimension 5 is equiangular but is also not a SIC. In particular the trace product relation of a SIC is not satisfied. For the equiangular IC-POVM in dimension 7, we have a similar result.

Blake Stacey in reply to Planat Apr 12 2017 18:20 UTC

This is why I am confused (it is probably just a reading comprehension error on my part): If the POVM is IC, it must have at least $d^2$ elements. If it is a minimal IC-POVM, it must have exactly $d^2$ elements. But if it is minimal, IC and equiangular, then the angle is fixed by the requirement that the elements sum to the identity. Suppose that the trace of $\Pi_i \Pi_j$ is $\alpha$ whenever $i \neq j$. Summing this over all $j$ yields $1 + (d^2-1)\alpha$. But the projectors $\Pi_i$ themselves must sum to $dI$, so the value of $\alpha$ is fixed to $1/(d+1)$.

Planat Apr 13 2017 07:14 UTC

The trace of pairwise product of (distinct) projectors is not constant. For example, with the state $(0,1,-1,-1,1)$, one gets an equiangular IC-POVM in which the trace is trivalued: it is either $1/16$, or $(7 \pm 3\sqrt{5})/32$. For the state (0,1,i,-i,-1), there are five values of the trace.
We should explicit this observation in the next version of the paper.

Robin Blume-Kohout in reply to Planat Apr 13 2017 15:40 UTC (1 points)

This appears to be an odd and nonstandard definition of "equiangular", unless I'm missing something? Most references I'm aware of, including [Wikipedia][1] and [Renes et al 2004][2] agree that "a set of lines is called equiangular if every pair of lines makes the same angle". For unit vectors (rays), that means they have the same inner product (or absolute value thereof).

In the case you're describing, it sounds like the angles are not all equal, but chosen from a small discrete set, as (e.g.) in [this recent paper][3], or some of the older references therein. MUBs and stabilizer states are examples of k-angle tight frames, albeit not minimal ones.


Planat Apr 13 2017 16:49 UTC

To define the complex angle, we used the (cyclotomic) field norm to the power one over the degree of the field, as stated in the introduction. It recovers the particular case of angles for SICs. In this sense "equiangular" means that all pairs of distinct lines make the same angle.

Marcus Appleby Apr 13 2017 18:54 UTC (2 points)

I am a little confused by this. As I use the term, lines are equiangular if and only if the "trace of pairwise product of (distinct) projectors is constant". You seem to be using the word in a different sense. It might be helpful if you were to explain exactly what is that sense.

Marcus Appleby Apr 13 2017 19:16 UTC (2 points)

I worded that badly, since you clearly have explained the sense in which you are using the word. I am wondering, however, how your definition relates to the usual one. Is it a generalization? Or just plain different? For instance, would a SIC be equiangular relative to your definition (using SIC field norm instead of cyclotomic field norm)?

Steve Flammia Apr 13 2017 19:16 UTC (2 points)

Just to clarify Michel's earlier remark, the field norm for the cyclotomics defines the norm in which these vectors are equiangular, and then they will generally **not** be equiangular in the standard norm based on the Hilbert-Schmidt inner product. In the example that he quotes,
$$\|(7\pm 3 \sqrt{5})/32\|_{\mathbb{Q}(\sqrt{5})} = \left[\frac{(7\pm 3 \sqrt{5})}{32} \frac{(7\mp 3 \sqrt{5})}{32}\right]^{1/\deg(\mathbb{Q}[\sqrt{5})]} = \frac{1}{16}.$$
It might be helpful in v2 of the paper if these vectors are called "generalized equiangular" or "equiangular with respect to the field norm", as this will help avoid confusion.

Edited Apr 13 2017 19:19 UTC by Steve Flammia

Blake Stacey in reply to Steve Flammia Apr 14 2017 00:35 UTC

I agree with Steve Flammia's comment. The field norm is a nice generalization of the standard norm. (I haven't yet thought about whether there might be a physics motivation for it, rather than a purely mathematical one, but that's not important right now.) To avoid confusion, some phrase like "equiangular with respect to the field norm" or "field-norm equiangular" should be used.

Robin Blume-Kohout Apr 14 2017 03:03 UTC

Okay, I see the resolution to my confusion now (and admit that I was confused). Thanks to Michel, Marcus, Blake, and Steve!

Since I don't know the first thing about cyclotomic field norms... can anybody explain the utility of this norm, for this problem? I mean, just to be extreme, I could define a trivial norm that is 1 for all vectors except $\vec{0}$, and then all rank-1 POVMs would be equiangular. I'm not by any means saying that this is what's done here! My point is that there exist norms for which equi-angularity is less interesting than others. The Hilbert-Schmidt norm is very relevant for quantum states in Hilbert spaces, because it's what appears in Born's rule. What can I do with this field norm that makes it interesting and relevant?

P.S. @Marcus, if I'm understanding this correctly, then whenever two pairs have equal Hilbert-Schmidt norm, they will have equal field norm (but different H-S norms can correspond to equal field norms). So SICs should still be equiangular in field norm. Unless I'm misunderstanding again!

Edited Apr 14 2017 03:07 UTC by Robin Blume-Kohout

Planat Apr 14 2017 08:11 UTC (1 points)

First of all, thanks to all for helping to clarify some hidden points of our paper.
As you can see, the field norm generalizes the standard Hilbert-Schmidt norm.
It works for SIC [e.g. d=2, d=3 (the Hesse) and d=8 (the Hoggar)].

The first non-trivial case is with d=4 when one needs to extend the rational field
by a 12th root of unity, i.e. n=GCD(d,r)=GCD(4,3)=12, that is r=3 for defining the appropriate
fiducial state and d=4 to allow the action of the two-qubit Pauli group on it.
Then one needs the field norm in the so defined cyclotomic extension to normalize the vectors of the
resulting IC-POVM. This IC is dichotomic in angles and traces of paiwise products.

Incidently, such a 4-dimensional IC is related to the Mermin square through the traces of triple products.