A Sequence of Qubit-Qudit Pauli Groups as a Nested

Structure of Doilies
Metod Saniga1 and Michel Planat2
1

arXiv:1102.3281v2 [quant-ph] 12 Apr 2011

Astronomical Institute, Slovak Academy of Sciences

SK-05960 Tatransk
a Lomnica, Slovak Republic
([email protected])
2
Institut FEMTO-ST, CNRS, 32 Avenue de lObservatoire
F-25044 Besancon Cedex, France
([email protected])
(12 April 2011)

Abstract
Following the spirit of a recent work of one of the authors (J. Phys. A: Math. Theor. 44 (2011)
045301), the essential structure of the generalized Pauli group of a qubit-qudit, where d = 2k
and an integer k 2, is recast in the language of a finite geometry. A point of such geometry
is represented by the maximum set of mutually commuting elements of the group and two
distinct points are regarded as collinear if the corresponding sets have exactly 2k 1 elements
in common. The geometry comprises 2k 1 copies of the generalized quadrangle of order two
(the doily) that form 2k1 1 pencils arranged into a remarkable nested configuration. This
nested structure reflects the fact that maximum sets of mutually commuting elements are of
two different kinds (ordinary and exceptional) and exhibits an intriguing alternating pattern:
the subgeometry of the exceptional points of the (k + 2)-case is found to be isomorphic to the
full geometry of the k-case. It should be stressed, however, that these generic properties of the
qubit-qudit geometry were inferred from purely computer-handled cases of k = 2, 3, 4 and 5
only and, therefore, their rigorous, computer-free proof for k 6 still remains a mathematical
challenge.
PACS Numbers: 03.65.Aa 03.65.Fd 02.10.Ox
Keywords: Qubit-Qudit Generalized Pauli Group Generalized Quadrangle of Order Two
Nested Structure

Introduction

Although finite geometries/point-line incidence structures behind the generalized Pauli groups of
single qudits have essentially all been discovered and thoroughly explored [1][4], only little is
still known about those underlying the groups of multi-qudits and/or tensor products of qudits of
different ranks. A number of particular cases have been analyzed by computer [5][9], and certain
easy-to-handle configurations have also been studied and completely described in a computerfree fashion [4],[7], [10][16]. Based on the recent work of one of us [9], we shall address below a
particular case of the tensor product of a qubit and a single qudit, where d = 2k and an integer
k 2. Analyzing the first four cases in the sequence, viz. k = 2, 3, 4 and 5, by computer was
already sufficient for us to infer the generic pattern of geometries behind the corresponding Pauli
groups for any k 2. This sequence of finite geometries is characterized by a remarkable nesting of
fundamental building blocks that are nothing but generalized quadrangles of order two, GQ(2, 2)s.

This seems to be the most crucial finding given the facts that this generalized quadrangle 1) is the
geometry behind the generalized Pauli group of two-qubits [7, 11] and 2) also plays, together with
other prominent point-line incidence structures, an essential role in finite geometrical aspects of
the still mysterious black-hole/qubit correspondence [17, 18].

Qudits, Generalized Pauli Groups and the Doily

Given an integer d > 1 and Zd := {0, 1, . . . , d 1}, where addition and multiplication of elements
from Zd is understood modulo d, we consider the d-dimensional complex Hilbert space Cd and
denote by
{ |si : s Zd }
a computational basis of Cd . A qudit in Cd is a vector
|i =

d1
X

s |si, where s C and

s=0

d1
X

|s |2 = 1.

s=0

Taking (d) to be a fixed primitive d-th root of unity (e. g., (d) = exp(2i/d)), we define unitary
s
X(d) (shift) and Z(d) (clock) operators on Cd via X(d) |si = |s + 1i and Z(d) |si = (d)
|si for all
s Zd . In the computational basis

1 0 0 ...
0
0 0 ... 0 1
1 0 . . . 0 0
0 0 . . .
0

0 1 . . . 0 0
0 0 2 . . .
0
X(d) =
(1)
and Z(d) =
.
.. .. . .
.. ..
. .
.. . .
..
. .
. .
.
. .. ..
.
.
0

0 ... 1

. . . d1

The (generalized) Pauli group generated by X(d) and Z(d) will be denoted as G(d) . For all s Zd
s+1
s
we have X(d) Z(d) |si = (d)
|s + 1i and Z(d) X(d) |si = (d)
|s + 1i. This gives the basic relation
(d) X(d) Z(d) = Z(d) X(d)
which implies that each element of G(d) can be written in the unique normal form
a
b
c
(d)
X(d)
Z(d)
for some integers a, b, c Zd .

(2)

The uniqueness of this normal form implies that G(d) is a group of order d3 . From (2) it is also
readily seen that
b c+a+a b+b c+c
a
b
c
a
b
c
((d)
X(d)
Z(d)
)((d)
X(d)
Z(d)
) = (d)
X(d) Z(d) .
which implies that commutator of two operators W and W ,
[W, W ] := W W W 1 W

acquires in our case the form

cb c b
a
b
c
a
b
c
[(d)
X(d)
Z(d)
, (d)
X(d)
Z(d)
] = (d)
I(d) .

(3)

Let us recall that two operators commute if, and only if, their commutator (taken in any order)
is equal to I(d) (the identity matrix); hence, G(d) is a non-commutative group. We also mention
in passing that there are two important normal subgroups of G(d) : its centre Z(G(d) ) and its
commutator subgroup G(d) , the two being identical
a
G(d) = Z(G(d) ) = {(d)
I(d) : a Zd }.

It is important to observe that a and a do not occur on the right-hand side of eq. (3). So, in order
to study the commutation relations between the elements of G(d) one can disregard the complex
a
b
c
phase factors (d)
and work solely with d2 matrix products X(d)
Z(d)
.
The final notion we shall need in the sequel is that of a finite generalized quadrangle of order
(s, t), usually denoted GQ(s, t). This is an incidence structure S = (P, B, I), where P and B are
disjoint (non-empty) sets of objects, called respectively points and lines, and where I is a symmetric
point-line incidence relation satisfying the following axioms [19]: (i) each point is incident with
1+t lines (t 1) and two distinct points are incident with at most one line; (ii) each line is incident
with 1 + s points (s 1) and two distinct lines are incident with at most one point; and (iii) if x is
a point and L is a line not incident with x, then there exists a unique pair (y, M ) P B for which
xIM IyIL; from these axioms it readily follows that |P | = (s + 1)(st + 1) and |B| = (t + 1)(st + 1). If
s = t, S is said to have order s. A generalized quadrangle with both s > 1 and t > 1 is called thick.
The smallest thick generalized quadrangle is obviously the (unique) GQ(2, 2), often dubbed the
doily. This quadrangle is endowed with 15 points/lines, with each line containing three points
and, dually, each point being on three lines. Moreover, it is a self-dual object, i. e., isomorphic to its
dual. In the older literature this geometry is also known as the Cremona-Richmond configuration
(see, e. g., [20]) and essential features of its structure are depicted in Figure 1.

Figure 1: A picture of the generalized quadrangle of order two, the doily. The circles stand for its
points, while its lines are given by triples of points lying on the same segments (10) and/or arcs of
circles (5). (Note that this point-line incidence geometry does not contain any triangles.)

The Qubit-Qu2k it Pauli Group and Its Geometry

As already mentioned, we shall focus on the generalized Pauli group associated with the Hilbert
k
space of the form C2 C2 , where an integer k 2.
We shall first handle the k = 2 case. Disregarding the phase factors, from eqs. (1) and (2) one
sees that the elements of the group associated with C2 (qubit) are
{I(2) , X(2) , Z(2) , X(2) Z(2) },
and those for C4 read
2
3
2
3
2
{I(4) , X(4) , X(4)
, X(4)
, Z(4) , X(4) Z(4) , X(4)
Z(4) , X(4)
Z(4) , Z(4)
,
2
2
2
3
2
3
3
2
3
3
3
X(4) Z(4)
, X(4)
Z(4)
, X(4)
Z(4)
, Z(4)
, X(4) Z(4)
, X(4)
Z(4)
, X(4)
Z(4)
}.

From the last two equations it follows that the generalized Pauli group of C2 C4 will feature the
following elements (again disregarding the phase factors)
3
3
3
2
2
,
Z(4)
, I(2) X(4)
Z(4)
, . . . , I(2) X(4)
{I(2) I(4) , I(2) X(4) , I(2) X(4)
3
3
3
2
2
,
Z(4)
, X(2) X(4)
Z(4)
, . . . , X(2) X(4)
X(2) I(4) , X(2) X(4) , X(2) X(4)
3
3
3
2
2
,
Z(4)
, Z(2) X(4)
Z(4)
, . . . , Z(2) X(4)
Z(2) I(4) , Z(2) X(4) , Z(2) X(4)
3
3
2
}.
Z(4)
, . . . , X(2) Z(2) X(4)
X(2) Z(2) I(4) , X(2) Z(2) X(4) , X(2) Z(2) X(4)

Since I(2) I(4) commutes with every other element, we consider only the remaining 63 elements/operators and, for convenience, number them from 1 to 63 in the order as they appear
in the last equation.
To find the geometry behind this Pauli group, we first look for maximum sets of mutually
commuting elements. We find that each such set features seven elements and their total number
is 39. Three of them, namely
a {2, 8, 10, 16, 18, 24, 26},
b {2, 8, 10, 32, 34, 40, 42},
c {2, 8, 10, 48, 50, 56, 58},
are special (and henceforth referred to as exceptional) in the sense that each of them shares with
any other set at least one element. The remaining 36 sets split into three distinct, equally-sized
families, in particular family I
1
2

{1, 2, 3, 16, 17, 18, 19},

{1, 2, 3, 32, 33, 34, 35},

3
4

{1, 2, 3, 48, 49, 50, 51},

{2, 9, 11, 16, 18, 25, 27},

5
6

{2, 9, 11, 32, 34, 41, 43},

{2, 9, 11, 48, 50, 57, 59},

7
8

{2, 24, 26, 33, 35, 57, 59},

{2, 24, 26, 41, 43, 49, 51},

9
10

{2, 25, 27, 33, 35, 56, 58},

{2, 25, 27, 40, 42, 49, 51},

11
12

{2, 17, 19, 40, 42, 57, 59},

{2, 17, 19, 41, 43, 56, 58},

1
2

{4, 8, 12, 16, 20, 24, 28},

{4, 8, 12, 32, 36, 40, 44},

3
4

{4, 8, 12, 48, 52, 56, 60},

{6, 8, 14, 16, 22, 24, 30},

5
6

{6, 8, 14, 32, 38, 40, 46},

{6, 8, 14, 48, 54, 56, 62},

{8, 20, 28, 34, 42, 54, 62},

8
9

{8, 20, 28, 38, 46, 50, 58},

{8, 22, 30, 34, 42, 52, 60},

10
11

{8, 22, 30, 36, 44, 50, 58},

{8, 18, 26, 36, 44, 54, 62},

12

{8, 18, 26, 38, 46, 52, 60},

family II

and family III

1

{5, 10, 15, 16, 21, 26, 31},

2
3

{5, 10, 15, 32, 37, 42, 47},

{5, 10, 15, 48, 53, 58, 63},

4
5

{7, 10, 13, 16, 23, 26, 29},

{7, 10, 13, 32, 39, 42, 45},

6
7

{7, 10, 13, 48, 55, 58, 61},

{10, 21, 31, 34, 40, 55, 61},

8
9

{10, 21, 31, 39, 45, 50, 56},

{10, 23, 29, 34, 40, 53, 63},

10

{10, 23, 29, 37, 47, 50, 56},

11
12

{10, 18, 24, 37, 47, 55, 61},

{10, 18, 24, 39, 45, 53, 63}.

It is easy to verify that this split into three above-given families is unique as the twelve sets in each
2
2
2
2
family have a single element in common (2 = I(2) X(4)
, 8 = I(2) Z(4)
and 10 = I(2) X(4)
Z(4)
,
respectively).
Next, we define the point-line incidence geometry whose points are these 39 maximum sets of
mutually commuting elements and where two points are collinear if the corresponding sets have
exactly three (22 1) elements in common. The three exceptional sets then form a distinguished line
in this geometry and each family generates a copy of the doily from which one line is deleted; this
line being nothing but the distinguished line! Hence, this point-line incidence structure comprises
three doilies on a common line, a pencil of doilies as illustrated in Figure 2.

6*

12*
5*

3*

11*

9*

10*

2*

8
5

11

10

12

4*

8
5

10

12

11

b
1*

8*

7*

Figure 2: A diagrammatical illustration of the finite geometry behind the qubit-ququartit Pauli
group: a set of three generalized quadrangles of order two sharing a line.
In a completely analogous way, we shall analyze the k = 3 case. Here, the relevant elements
of the Pauli group related to the Hilbert space C8 are (after dropping the subscript 8 for better
readability of the formulas)
{I, X, X 2, X 3 , X 4 , X 5 , X 6 , X 7 , Z, ZX, ZX 2, ZX 3 , ZX 4 , ZX 5 , ZX 6 , ZX 7 , Z 2 , Z 2 X,
Z 2 X 2 , Z 2 X 3 , Z 2 X 4 , Z 2 X 5 , Z 2 X 6 , Z 2 X 7 , Z 3 , Z 3 X, Z 3 X 2 , Z 3 X 3 , Z 3 X 4 , Z 3 X 5 ,
Z 3 X 6 , Z 3 X 7 , Z 4 , Z 4 X, Z 4 X 2 , Z 4 X 3 , Z 4 X 4 , Z 4 X 5 , Z 4 X 6 , Z 4 X 7 , Z 5 , Z 5 X, Z 5 X 2 ,
Z 5 X 3 , Z 5 X 4 , Z 5 X 5 , Z 5 X 6 , Z 5 X 7 , Z 6 , Z 6 X, Z 6 X 2 , Z 6 X 3 , Z 6 X 4 , Z 6 X 5 , Z 6 X 6 , Z 6 X 7 ,
Z 7 , Z 7 X, Z 7 X 2 , Z 7 X 3 , Z 7 X 4 , Z 7 X 5 , Z 7 X 6 , Z 7 X 7 },
which yield the following 255 (disregarding the identity) distinct matrices of the group associated
5

with C2 C8
{I(2) X, I(2) X 2 , . . . , I(2) Z 7 X 7 , X(2) I, X(2) X, . . . , X(2) Z 7 X 7 ,
Z(2) I, Z(2) X, . . . , Z(2) Z 7 X 7 , X(2) Z(2) I, . . . , X(2) Z(2) Z 7 X 7 }.
Again, to facilitate our reasoning, we number these elements from 1 to 255 consecutively as they
are listed in the last expression. Using computer, we find out that they form 87 maximum sets
of mutually commuting guys, each of cardinality 15. Fifteen of them are exceptional and they
explicitly read
a

{2, 4, 6, 32, 34, 36, 38, 64, 66, 68, 70, 96, 98, 100, 102},

b
c

{2, 4, 6, 32, 34, 36, 38, 128, 130, 132, 134, 160, 162, 164, 166},
{2, 4, 6, 32, 34, 36, 38, 192, 194, 196, 198, 224, 226, 228, 230},

d
e

{4, 16, 20, 32, 36, 48, 52, 64, 68, 80, 84, 96, 100, 112, 116},
{4, 16, 20, 32, 36, 48, 52, 128, 132, 144, 148, 160, 164, 176, 180},

f
g

{4, 16, 20, 32, 36, 48, 52, 192, 196, 208, 212, 224, 228, 240, 244},
{4, 18, 22, 32, 36, 50, 54, 64, 68, 82, 86, 96, 100, 114, 118},

h
i

{4, 18, 22, 32, 36, 50, 54, 128, 132, 146, 150, 160, 164, 178, 182},
{4, 18, 22, 32, 36, 50, 54, 192, 196, 210, 214, 224, 228, 242, 246},

j
k

{4, 32, 36, 80, 84, 112, 116, 130, 134, 162, 166, 210, 214, 242, 246},
{4, 32, 36, 80, 84, 112, 116, 146, 150, 178, 182, 194, 198, 226, 230},

l
m

{4, 32, 36, 82, 86, 114, 118, 130, 134, 162, 166, 208, 212, 240, 244},
{4, 32, 36, 82, 86, 114, 118, 144, 148, 176, 180, 194, 198, 226, 230},

n
o

{4, 32, 36, 66, 70, 98, 102, 144, 148, 176, 180, 210, 214, 242, 246},
{4, 32, 36, 66, 70, 98, 102, 146, 150, 178, 182, 208, 212, 240, 244}.

The remaining 72 ordinary sets split, as expected, into six sets of cardinality twelve each. These
six sets are found to form three pairs, namely pair A
1

{1, 2, 3, 4, 5, 6, 7, 64, 65, 66, 67, 68, 69, 70, 71},

2
3

{1, 2, 3, 4, 5, 6, 7, 128, 129, 130, 131, 132, 133, 134, 135},

{1, 2, 3, 4, 5, 6, 7, 192, 193, 194, 195, 196, 197, 198, 199},

{2, 4, 6, 33, 35, 37, 39, 64, 66, 68, 70, 97, 99, 101, 103},

5
6

{2, 4, 6, 33, 35, 37, 39, 128, 130, 132, 134, 161, 163, 165, 167},
{2, 4, 6, 33, 35, 37, 39, 192, 194, 196, 198, 225, 227, 229, 231},

7
8

{2, 4, 6, 96, 98, 100, 102, 129, 131, 133, 135, 225, 227, 229, 231},
{2, 4, 6, 96, 98, 100, 102, 161, 163, 165, 167, 193, 195, 197, 199},

9 {2, 4, 6, 97, 99, 101, 103, 129, 131, 133, 135, 224, 226, 228, 230},
10 {2, 4, 6, 97, 99, 101, 103, 160, 162, 164, 166, 193, 195, 197, 199},
11 {2, 4, 6, 65, 67, 69, 71, 160, 162, 164, 166, 225, 227, 229, 231},
12 {2, 4, 6, 65, 67, 69, 71, 161, 163, 165, 167, 224, 226, 228, 230},

1
2

{4, 17, 21, 34, 38, 51, 55, 64, 68, 81, 85, 98, 102, 115, 119},
{4, 17, 21, 34, 38, 51, 55, 128, 132, 145, 149, 162, 166, 179, 183},

3
4

{4, 17, 21, 34, 38, 51, 55, 192, 196, 209, 213, 226, 230, 243, 247},
{4, 19, 23, 34, 38, 49, 53, 64, 68, 83, 87, 98, 102, 113, 117},

5
6

{4, 19, 23, 34, 38, 49, 53, 128, 132, 147, 151, 162, 166, 177, 181},
{4, 19, 23, 34, 38, 49, 53, 192, 196, 211, 215, 226, 230, 241, 245},

7
8

{4, 34, 38, 81, 85, 115, 119, 130, 134, 160, 164, 211, 215, 241, 245},
{4, 34, 38, 81, 85, 115, 119, 147, 151, 177, 181, 194, 198, 224, 228},

{4, 34, 38, 83, 87, 113, 117, 130, 134, 160, 164, 209, 213, 243, 247},

10
11

{4, 34, 38, 83, 87, 113, 117, 145, 149, 179, 183, 194, 198, 224, 228},
{4, 34, 38, 66, 70, 96, 100, 145, 149, 179, 183, 211, 215, 241, 245},

12

{4, 34, 38, 66, 70, 96, 100, 147, 151, 177, 181, 209, 213, 243, 247},

1
2

{8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88, 96, 104, 112, 120},
{8, 16, 24, 32, 40, 48, 56, 128, 136, 144, 152, 160, 168, 176, 184},

3
4

{8, 16, 24, 32, 40, 48, 56, 192, 200, 208, 216, 224, 232, 240, 248},
{12, 16, 28, 32, 44, 48, 60, 128, 140, 144, 156, 160, 172, 176, 188},

pair B

{12, 16, 28, 32, 44, 48, 60, 192, 204, 208, 220, 224, 236, 240, 252},

6
7

{12, 16, 28, 32, 44, 48, 60, 64, 76, 80, 92, 96, 108, 112, 124},
{16, 32, 48, 72, 88, 104, 120, 132, 148, 164, 180, 204, 220, 236, 252},

8
9

{16, 32, 48, 72, 88, 104, 120, 140, 156, 172, 188, 196, 212, 228, 244},
{16, 32, 48, 76, 92, 108, 124, 132, 148, 164, 180, 200, 216, 232, 248},

10
11

{16, 32, 48, 76, 92, 108, 124, 136, 152, 168, 184, 196, 212, 228, 244},
{16, 32, 48, 68, 84, 100, 116, 140, 156, 172, 188, 200, 216, 232, 248},

12

{16, 32, 48, 68, 84, 100, 116, 136, 152, 168, 184, 204, 220, 236, 252},

{10, 20, 30, 32, 42, 52, 62, 128, 138, 148, 158, 160, 170, 180, 190},

2
3

{10, 20, 30, 32, 42, 52, 62, 192, 202, 212, 222, 224, 234, 244, 254},
{14, 20, 26, 32, 46, 52, 58, 64, 78, 84, 90, 96, 110, 116, 122},

{14, 20, 26, 32, 46, 52, 58, 128, 142, 148, 154, 160, 174, 180, 186},

5
6

{14, 20, 26, 32, 46, 52, 58, 192, 206, 212, 218, 224, 238, 244, 250},
{10, 20, 30, 32, 42, 52, 62, 64, 74, 84, 94, 96, 106, 116, 126},

7
8

{20, 32, 52, 74, 94, 106, 126, 132, 144, 164, 176, 206, 218, 238, 250},
{20, 32, 52, 74, 94, 106, 126, 142, 154, 174, 186, 196, 208, 228, 240},

9 {20, 32, 52, 78, 90, 110, 122, 132, 144, 164, 176, 202, 222, 234, 254},
10 {20, 32, 52, 78, 90, 110, 122, 138, 158, 170, 190, 196, 208, 228, 240},
11 {20, 32, 52, 68, 80, 100, 112, 138, 158, 170, 190, 206, 218, 238, 250},
12 {20, 32, 52, 68, 80, 100, 112, 142, 154, 174, 186, 202, 222, 234, 254},

and pair C
1

{9, 18, 27, 36, 45, 54, 63, 64, 73, 82, 91, 100, 109, 118, 127},

2
3

{9, 18, 27, 36, 45, 54, 63, 128, 137, 146, 155, 164, 173, 182, 191},
{9, 18, 27, 36, 45, 54, 63, 192, 201, 210, 219, 228, 237, 246, 255},

4
5

{18, 36, 54, 73, 91, 109, 127, 132, 150, 160, 178, 205, 223, 233, 251},
{18, 36, 54, 73, 91, 109, 127, 141, 159, 169, 187, 196, 214, 224, 242},

{18, 36, 54, 77, 95, 105, 123, 132, 150, 160, 178, 201, 219, 237, 255},

7
8

{18, 36, 54, 77, 95, 105, 123, 137, 155, 173, 191, 196, 214, 224, 242},
{18, 36, 54, 68, 86, 96, 114, 137, 155, 173, 191, 205, 223, 233, 251},

9
10

{13, 18, 31, 36, 41, 54, 59, 64, 77, 82, 95, 100, 105, 118, 123},
{13, 18, 31, 36, 41, 54, 59, 128, 141, 146, 159, 164, 169, 182, 187},

11
12

{13, 18, 31, 36, 41, 54, 59, 192, 205, 210, 223, 228, 233, 246, 251},
{18, 36, 54, 68, 86, 96, 114, 141, 159, 169, 187, 201, 219, 237, 255},

e
1 {11, 22, 25, 36, 47, 50, 61, 64, 75, 86, 89, 100, 111, 114, 125},
e
2 {11, 22, 25, 36, 47, 50, 61, 128, 139, 150, 153, 164, 175, 178, 189},
e
3 {11, 22, 25, 36, 47, 50, 61, 192, 203, 214, 217, 228, 239, 242, 253},
e
4 {22, 36, 50, 75, 89, 111, 125, 132, 146, 160, 182, 207, 221, 235, 249},
e
5 {22, 36, 50, 75, 89, 111, 125, 143, 157, 171, 185, 196, 210, 224, 246},

e
6 {22, 36, 50, 79, 93, 107, 121, 132, 146, 160, 182, 203, 217, 239, 253},
e
7 {22, 36, 50, 79, 93, 107, 121, 139, 153, 175, 189, 196, 210, 224, 246},
e
8 {15, 22, 29, 36, 43, 50, 57, 64, 79, 86, 93, 100, 107, 114, 121},
e
9 {15, 22, 29, 36, 43, 50, 57, 128, 143, 150, 157, 164, 171, 178, 185},

f
10 {15, 22, 29, 36, 43, 50, 57, 192, 207, 214, 221, 228, 235, 242, 249},
f
11 {22, 36, 50, 68, 82, 96, 118, 139, 153, 175, 189, 207, 221, 235, 249},
f {22, 36, 50, 68, 82, 96, 118, 143, 157, 171, 185, 203, 217, 239, 253}.
12

One observes that each twelve-element set features three common elements and the two sets
4
4
in a pair share a single element, this being the element 4 = I(2) X(8)
, 32 = I(2) Z(8)
and
4
4
36 = I(2) Z(8) X(8) , respectively. Paralleling the preceding case, we again form the point-line
incidence geometry where points are the 87 maximum sets, but where two points are now collinear
if the corresponding sets have exactly seven (=23 1) elements in common. This geometry, depicted
in Figure 3, consists of seven doilies. These form three pencils, each comprising the doily generated
by the 15 exceptional sets and two doilies coming from pair A if the missing line {a, b, c} is added,
from pair B if the missing line {d, e, f } is added and from pair C if the missing line {g, h, i} is
added. The three lines in questions are pairwise skew; for the reader well-acquainted with the
structure of the doily it may be interesting to learn that in the dual of the distinguished doily
these lines answer to a tricentric triad of points.
For the next case in the hierarchy, k = 4 (qubit-quhexadecit), our computer calculations showed
that 1023 elements of the corresponding Pauli group form 183 maximum sets of 31 elements each,
out of which 39 are exceptional. We shall not go into much detail here but mention only the
outcome of our analysis, illustrated in line three of Figure 4. The geometry is defined as in the
previous two cases, save for the fact that collinearity is now synonymous with sharing fifteen (=
24 1) elements. The three central doilies in a pencil answer to the 39 elements of the exceptional
set, with the carrier line accounting for the finer, 36+3 split; hence, this subgeometry is isomorphic
to the full k = 2 case (see Figure 4). The 12 doilies at sides, underlying 144 ordinary points, form
8

12

12
9

8
7

10

10
2

7
9

12
12*

2*
3*

8*

5*

4*

1*

9*

11

11*

m
n

10

o
6*

7*

10*

11

11
3

j
g

~
11

~
10

9
~

c
2

10

11

8
12

~
12

Figure 3: A schematic sketch of the geometry behind the qubit-quoctic generalized Pauli group.
The seven doilies lie in three pencils on the distinguished doily (middle); the carrier lines of these
pencils are shown in boldface.

Figure 4: A compact illustration of the properties of the above-described finite geometries for (top
to bottom) k = 2, 3, 4 and 5. Each circle represents a doily and each triangle stands for a pencil of
doilies; the filled circle means that the corresponding doily contains solely exceptional points.
three quadruples, each being a satellite to one central doily. A quadruple further splits into two
pairs, and the two doilies in each pair form a pencil with the corresponding central doily. So we
have altogether seven pencils of doilies.
In the k = 5 case, we find 4095 elements of the associated Pauli group forming 375 maximum
sets of 63 elements each. The 87 of them are exceptional, and their geometry is fully isomorphic
to the k = 3 one (see Figure 4). The 24 satellite doilies account for the remaining 288 ordinary
points. They form 12 pairs and with six central doilies define 12 pencils; taking into account
three central pencils, this case is thus seen to be altogether endowed with 15 pencils.
These four cases obviously suffice to infer the geometry for an arbitrary k 2. This geometry,
which features 2k2 36 ordinary points, will comprise 2k 1 copies of the generalized quadrangle of
order two that form 2k1 1 pencils arranged into a remarkable nested configuration. The core of
this configuration consists of exceptional points and is isomorphic to the full k 2 geometry. From
Figure 4 it can readily be discerned how to construct generically the k + 2 geometry around the
k one. One simply takes a peripheral doily, i. e. any doily that does not consist of exceptional

10

points only, and associate with it two pairs of doilies in such manner that they form two pencils.
Obviously, we shall get two distinct sequence patterns according as the initial geometry is that of
the k = 2 or k = 3 case.

Conclusion

Following and extending the strategy adopted in [9], a finite geometrical treatment of the generalk
ized Pauli groups associated with the Hilbert spaces of type C2 C2 , where an integer k 2, has
been performed. A point-line incidence structure was defined as follows: its points are maximum
sets of pairwise commuting group elements and two distinct points are collinear if the corresponding sets share 2k 1 elements. The points were found to be of two distinct kinds, referred to as
ordinary and exceptional; a point of the latter kind being represented by such maximum set that
shares with any other maximum set at least one element. A computer-based analysis of the first
four cases in the sequence implies that, in general, this geometry features 2k 1 copies of the doily
that form 2k1 1 pencils arranged into a remarkable nested configuration reminding a fractal-like
behaviour with growing k. The core of this geometry, being generated by exceptional points, is
isomorphic to the full geometry of the k 2 case. Finding the doily, the smallest triangle-free
v3 -configuration, to be the fundamental building block of these geometries is pleasing of its own,
all the more that as already stressed in introduction this generalized quadrangle is the geometry behind the generalized Pauli group of two-qubits and it plays also an essential role in finite
geometrical aspects of the still mysterious black-hole/qubit correspondence. We hope that all the
above-discussed properties will soon be given a rigorous, computer-free proof.

Acknowledgements
This work was partially supported by the VEGA grant agency projects 2/0092/09 and 2/0098/10.
We are extremely grateful to Dr. Petr Pracna for providing us with improved electronic versions
of all the figures.

References
[1] Havlicek H, and Saniga M. Projective Ring Line of a Specific Qudit. J Phys A: Math Theor
2007;40:F943F952 (arXiv:0708.4333).
[2] Havlicek H, and Saniga M. Projective Ring Line of an Arbitrary Single Qudit. J Phys A: Math
Theor 2008;41:015302 (arXiv:0710.0941).
[3] Albouy O. The Isotropic Lines of Zd2 . J Phys A: Math Theor 2009;42:072001 (arXiv:0809.3220).
[4] Thas K. The Geometry of Generalized Pauli Operators of N-Qudit Hilbert Space, and an
Application to MUBs. EPL 2009;86:60005.
[5] Planat M, and Baboin A-C. Qudits of Composite Dimension, Mutually Unbiased Bases and
Projective Ring Geometry. J Phys A: Math Theor 2007;40:F1005F1012 (arXiv:0709.2623).
[6] Planat M, Baboin A-C, and Saniga M. Multi-Line Geometry of Qubit-Qutrit and Higher-Order
Pauli Operators. Int J Theor Phys 2008;47:11271135 (arXiv:0705.2538).
[7] Planat M, and Saniga M. On the Pauli Graphs of N-qudits. Quant Inform Comput 2008;8:127
146 (arXiv:quant-ph/0701211).
[8] Planat M. On the Geometry and Invariants of Qubits, Quartits and Octits. Int J Geom
Methods Mod Phys 2011;8, in press (arXiv:1005.1997).
[9] Planat M. Pauli Graphs When the Hilbert Space Dimension Contains a Square: Why the
Dedekind Psi Function? J Phys A: Math Theor 2011;44:045301 (arXiv:1009.3858).

11

[10] Saniga M, and Planat M. Multiple Qubits as Symplectic Polar Spaces of Order Two, Adv
Studies Theor Phys 2007;1:14 (arXiv:quant-ph/0612179).
[11] Saniga M, Planat M, and Pracna P. Projective Ring Line Encompassing Two-Qubits. Theor
Math Phys 2008;155:905913 (arXiv:quant-ph/0611063).
[12] Havlicek H, Odehnal B, and Saniga M. Factor-Group-Generated Polar Spaces and (Multi)Qudits. SIGMA 2009;5:096 (arXiv:0903.5418).
[13] Sengupta AR. Finite Geometries with Qubit Operators. Infin Dimens Anal Quant Probab
Relat Top 2009;12:359-366 (arXiv:0904.2812).
[14] Rau ARP. Mapping Two-Qubit Operators onto Projective Geometries. Phys Rev A
2009;79:42323 (arXiv:0808.0598).
[15] Havlicek H, Odehnal B, and Saniga M. Moebius Pairs of Simplices and Commuting Pauli
Operators. Math Pannonica 2010;21:115128 (arXiv:0905.4648).
[16] Vrana P, and Levay P. The Veldkamp Space of Multiple Qubits. J Phys A: Math Theor
2010;43:125303 (arXiv:0906.3655).
[17] Levay P, Saniga M, Vrana P, and Pracna P. Black Hole Entropy and Finite Geometry. Phys
Rev D 2009;79:084036 (arXiv:0903.0541).
[18] Levay P, Saniga M, Vrana P. Three-Qubit Operators, the Split Cayley Hexagon of Order Two
and Black Holes. Phys Rev D 2008;78:24022 (arXiv:0808.3849).
[19] Payne SE, and Thas JA. Finite Generalized Quadrangles. Pitman: Boston London Melbourne; 1984; see also Thas K. Symmetry in Finite Generalized Quadrangles. Birkhauser:
Basel; 2004.
[20] Coxeter HSM. Self-Dual Configurations and Regular Graphs. Bull Amer Math Soc
1950;56:413455.

12