Noah Snyder: Research Statement

Overview

Many objects and systems studied in science and mathematics are highly symmetric.
For example, the laws of nature don’t change if you translate everything in some direction,
the equations describing the state of an atom don’t change if you rotate around the nu-
cleus, and crystals have many translational, reflectional, and rotational symmetries. The
mathematical theory used to describe symmetry is called group theory, where a group
is a collection of symmetries together with a multiplication-like operation which describes
what happens when you apply first one symmetry and then another symmetry to produce
a third symmetry. Group theory has been developed intensely over the past two centuries,
and is one of the main subjects in the broad mathematical subject of algebra. One es-
pecially powerful application of group theory was given by Emmy Noether in 1918, who
proved that to any continuous family of symmetries of a physical system there is a corre-
sponding conservation law. Under Noether’s correspondence, translational symmetry in
space corresponds to conservation of momentum, rotational symmetry to conservation of
angular momentum, and time-translation symmetry to conservation of energy.
Although group theory is widely applicable and spectacularly successful, in some quan-
tum settings groups are no longer adequate to capture all symmetries. To capture these
more general symmetries, in the 1980s mathematicians developed theories of quantum
groups and tensor categories. Although these were originally discovered purely as math-
ematical objects, it has since become clear that certain tensor categories describe actual
physical materials in condensed matter physics called topological orders. These systems
are of very active interest to physicists and also to computer scientists who hope to use
them to build a quantum computer. Just as groups appear widely across mathematics, so
do tensor categories. As a result, my research touches on very different mathematical sub-
jects like operator algebras, representation theory, topology, category theory, and mathe-
matical physics, but in all cases my main objects of study are tensor categories.
More specifically, I have four main research programs: the classification of small in-
dex subfactors and small tensor categories, understanding exceptional tensor categories
via their quantum subgroups, the classification of simple skein theories, and using
topological quantum field theory and the cobordism hypothesis to better under-
stand the structure of tensor categories. The first three of these are very example-driven
and involve studying particular unusual mathematical objects and their properties, while
the fourth is more theory-driven. These programs also interact with different subjects of
mathematics: the first two of these programs are at the intersection of operator algebras
and representation theory, while the third is inspired by knot theory, and the fourth has
a heavily topological flavor. The first of these programs is largely complete (though I am
still doing some related work). The second program has achieved its main goals, but we
are still finishing a few papers giving some applications. The final two are in earlier stages
and are my main current areas of activity. I will conclude with a brief remark on the rela-
tionship between my work and physics and computer science.
1
 1. Classification of Small Index Subfactors

One of the first spectacular applications of group theory was Galois’s study of inclusions
of fields in 1832, where he used group theory to explain why there is a formula for solving
quadratic equations, but no formula for solving fifth degree equations. Similarly, one of the
first appearances of quantum groups, was as Galois-like symmetry group for inclusions of
von Neumann factors. In particular, in the early 1980s Vaughan Jones defined an index
[N : M ] which measures the size of such a subfactor, and proved a stunning result that be-
low 4 the only possible values for the index are numbers of the form 4(cos πn )2 for a natural
number n [Jon83]. The techniques he used in his proof led to the Jones polynomial in knot
theory [Jon85]. In 1990, Jones received the Fields Medal, the highest honor in Mathemat-
ics, for this work on subfactors.
The Jones polynomial assigns to any framed knot (to a mathematician, this means you
tie a shoelace into a complicated knot and then fuse the ends together so that the knot
cannot be untied) a polynomial in A and A−1 which is invariant in the sense that if you do
any manipulation to the knot that doesn’t break it into pieces the associated polynomial
will not change. Following Kauffman [Kau87], the Jones polynomial can be computed by
starting with your knot, smashing it down into the plane, and applying the following two
rules. The second rule lets you replace all of the crossings to get a giant sum of diagrams
which each consist of a bunch of circles that are not knotted, and then the first rule lets
you turn those circles into a polynomial.
= −A2 − A−2

=A + A−1

The connection between subfactors and knot theory comes from an interpretation of the
“standard invariant” of a subfactor as a planar algebra, which is way of doing complex
calculations in tensor categories or subfactors by using skein theoretic relations like the
above two. The square of the value of the circle in the first equation is exactly the Jones
index of the corresponding subfactor.
Since the subfactor indices below 4 are discrete, it was natural to look for a classifica-
tion of subfactors of index less than 4, which was completed in work of Jones, Ocneanu,
Popa, and others in the early 1990s [Ocn88]. The main technique was to study subfactors
via something called the principal graph, which explains the induction/restriction rules for
turning N -N bimodules into N -M bimodules and vice-versa. Their results showed that
indeed there is a very nice classification of subfactors of index below 4, with two infinite
families (An and D2n ) and four additional examples (E6 , E6 , E8 , and E8 ). Each of these
gives an interesting example of a quantum group, in fact they are quantum analogues of
the symmetry groups of the Platonic solids. At index equal to 4 there is a similar classi-
fication in terms of affine Dynkin diagrams and some additional cohomological data, and
this classification is closely related now to the classical Platonic solids [Pop94, p. 231].
These classifications can either be stated purely algebraically in terms of something called
the “standard invariant,” or can be thought of analytically as a classification of subfactors
of a particular factor called the hyperfinite II1 factor via a deep theorem of Popa [Pop94].
 Figure 1. The initial steps of the jellyfish algorithm. The dotted ovals
represent linear combinations of Temperley-Lieb diagrams.

My first main research program was to extend the classification of subfactors from in-
dex 4 to index 5 in joint work with Morrison, Peters, and many others. This program
was started by Haagerup in 1994 [Haa94], who realized that once you exclude one con-
tinuous
√ family, the possible indices for subfactors again become discrete between 4 and
3 + 3 [Haa94]. My first major contribution to this field was the construction of the ex-
tended Haagerup subfactor in a paper that appeared in Acta Mathematica (one of the top
5 journals in mathematics) [BMPS12]. This
√ construction was the final step in the classifi-
cation of subfactors of index up to 3 + 3 that began 15 years earlier by Haagerup. The
construction uses Jones’s theory of planar algebras (which are a 2-dimensional picture cal-
culus for doing subfactor calculations) and followed an outline developed by Jones and by
Peters in her thesis [Pet10]. A key step was a new skein-theoretic “evaluation algorithm”
we called the Jellyfish algorithm (summarized in Figure 1). Our construction was com-
puter assisted, and a key point was to better understand the number theoretic properties
of this planar algebra in order to be able to do exact arithmetic in a reasonable amount of
computer time.
We then set our sights on getting to index 5, which was quite a lot harder and required
several theoretical improvements on Haagerup’s techniques, as well as developing computer
software to speed up the calculations. We eventually completed this classification in half-
a-dozen other papers, eventually showing that there are exactly five pairs of amenable sub-
factors between index 4 and 5. These papers appeared in high quality math and mathe-
matical physics journals. Furthermore, we wrote a survey article on this project which ap-
peared in the Bulletin of the American Mathematical Society, which is sent to all members
of the AMS and is the premier math journal for survey articles.
 The statement of the classification theorem is given below and represents the culmina-
tion of work of many authors. It is proved in the papers [MS12, MPPS12, IJMS12, PT12],
which in turn rely on constructions given in [AH99,BMPS12,GdlHJ89,Izu01], and a unique-
ness result from [Han10]. The combinatorial enumeration of possible principal graphs in
[MS12] builds on the earlier work of [Haa94]. The classification relies on the powerful num-
ber theoretic obstructions proved in [CMS11] which generalize√ earlier work of [AY09]. This
theorem supersedes an earlier classification below index 3 + 3 from [Haa94, Bis98, AH99,
AY09, BMPS12].
Theorem 1. If an extremal subfactor has index between 4 and 5 its standard invariant is
Temperley-Lieb (with principal graph A∞ ) or it is one of the following ten planar algebras.
√
5+ 13
• The Haagerup planar algebra [AH99], with index 2
and principal bigraph pair
 
,

and its dual.

8 2
q
13
√

• The extended Haagerup planar algebra [BMPS12], with index +
3 3
Re 3
2
−5 − 3i 3
and principal bigraph pair
 
,

and its dual.

√
5+ 17
• The Asaeda-Haagerup planar algebra [AH99], with index 2
and principal bi-
graph pair
 
,

and its dual.

√
• The 3311 Goodman-de la Harpe-Jones planar algebra [GdlHJ89], with index 3 + 3
and principal bigraph pair
 
,

and its dual (since it is not self-dual despite having the same principal and dual
principal graphs [Kaw95]).
• Izumi’s
√
self-dual 2221 planar algebra [Izu01] and its complex conjugate, with index
5+ 21
2
and principal bigraph pair
 
, .

This project has attracted further interest, and recently it was extended to index 5.25
in work of Afzaly–Morrison–Penneys [AMP15] based on breakthroughs√ of Penneys and
his collaborators [Pen15, BP14] and a key result at index 3 + 5 by Liu [Liu15]. In my
work I have focused in new directions, namely better understanding the exotic examples
which appear in this classification and working on a different skein theoretic classification
program. Nonetheless, there are still several small index classification questions that I am
still working on, namely, extending this classification to small index bimodules, proving an
analogous classification in the non-unitary setting of interest to representation theorists,
considering the A∞ subfactors excluded in Haagerup’s approach, and classifying quadrilat-
erals of small index subfactors.

2. Brauer–Picard groupoids and exceptional subfactors

One of the main goals in proving classification theorems is to find new unknown and
complicated examples. For example, the E8 Lie algebra was only found by Cartan–Killing’s
classification of all semisimple Lie algebras, and the Monster Group was found only by
classifying all finite simple groups. Similarly, our small index subfactor classification is es-
sentially the only way that “exceptional” or “exotic” quantum groups have been found.
So I have now turned my attention to better understanding these examples in my second
program which is joint with Grossman. We study the exotic small index subfactors via
their “quantum subgroups.” Another way of thinking about this is that we simultaneously
study all subfactors that can be built out of a fixed collection of bimodules. Our approach
is similar to Ocneanu’s notion of a “maximal atlas” [Ocn01] and to Etingof–Nikshych–
Ostrik’s Brauer–Picard groupoid [ENO10]. This approach has two main advantages, first
having a richer structure gives stronger constraints than you get studying candidate sub-
factors one at a time, and second one of the other subfactors appearing might be easier
to understand (though larger index). Grossman and I were the first people to apply this
approach to understanding exotic subfactors, and we have used it to answer an old open
problem.
We have applied this program to both the Haagerup and Asaeda–Haagerup subfac-
tors [GS12b, GS12a, GJS12, GIS15], and used these calculations to construct many new
subfactors and several new tensor categories. These papers all appeared in strong journals,
with the last paper appearing in the oldest mathematics journal Journal für die Reine und
Angewandte Mathematik (Crelle’s Journal). In the Haagerup case, there are exactly three
fusion categories in the Morita equivalence class and a unique Morita equivalence between
each of them. One of these fusion categories was new and very interesting. We also gave
a two line construction of Grossman–Izumi’s “Haagerup plus one” subfactor, which pre-
viously took 14 pages of complicated calculations. The Asaeda–Haagerup case proved to
be much more interesting. There are now six fusion categories (four of which are new)
and four Morita equivalences between each of them (most of which were new). Further-
more, several of these new fusion categories are more highly symmetric than the original
Asaeda–Haagerup subfactor, see Figure 2. This allowed us, in joint work with Grossman
and Izumi, to give a symmetric construction of the Asaeda-Haagerup subfactor. This new
construction has allowed us to answer many open questions about this exotic quantum
group, including one question that was open for 20 years. We are currently working on
several further applications of our construction, to extensions as in our Quantum Topol-
ogy paper, and to lattices of intermediate subfactors. Combined with some ideas of Izumi
and Evans–Gannon [Izu01, Izu, EG11], our construction also gives the first evidence that
Asaeda–Haagerup is related to an infinite family of quantum groups.
 11
00
00
11 0
1
0
1
000000
111111
000000 111111
111111 000000
000000
111111
000000
111111
000000
111111 000000
111111
000000
111111
1
0
000000
111111 00
11
000000
111111
000000
111111 000000
111111
000000
111111
00
11 000000
111111
0
1
000000
111111
00
11
0000
1111 000000
111111
00000
111110
1
000000
111111
0000
1111
000000
111111 000000
111111
00000
11111
000000
111111
0000
1111
000000
11111100000
11111
0000
1111 000000
111111
00000
11111
000000
111111 100
0
0000
111111
000000
111111
00000
11111
000000
111111
0000
1111
000000
11111100000
11111
0000
1111 000000
111111
00000
11111
000000
111111
00
11
000000
0000
1111
11111100000
111110
1
000000
111111
00
11
000000
111111 0
1
000000
111111
000000
111111
000000
111111 000000
111111
000000
111111
000000
111111
1
0 000000
111111
00
11
000000
111111
000000 111111
000000
111111
00
11
000000
111111
000000
111111
000000
1111110
1
00
11
000000 111111
111111 0000000
1

Figure 2. The principal graphs of the original Asaeda–Haagerup subfactor

and of our newly constructed subfactor Morita equivalent to it. Note that
the original only has two-fold symmetry, while ours has four-fold symmetry.

3. Classification of simple skein theories

With Morrison, Peters, and Thurston we have started a skein-theoretic classification

project which we hope will provide another source of exotic examples. Perhaps a helpful
analogy is that looking for new exceptional tensor categories is like looking for new kinds
of objects in the sky, you build one “telescope” (i.e. classification program) and then you
look at all objects in the some part of the sky which are bright enough to be seen by that
telescope (i.e. “small index” means brighter) and then you try to understand them. In or-
der to find more objects you could make the telescope stronger (i.e. increase the index), or
instead you could build a new kind of telescope at a different frequency (i.e. tensor cate-
gories with simple skein theories rather than small index).
Specifically we look at planar algebras which can be described via simple skein relations
analogous to knot polynomials, but using trivalent planar graphs instead of knots. A pla-
nar graph is just a collection of vertices connected by non-intersecting edges, and a graph
is called trivalent if three edges come in to each vertex. So an example of such a graph is
a flattened cube (with 8 vertices, 12 edges, and 6 faces). Just as skein theories for knots
involve relations imposed on small parts of knots, skein theories for planar trivalent graphs
involve relations that you can apply to a small part of a trivalent graph, like
1 1
− + − = 0.
d−1 d−1

Our first result about trivalent skein planar appears in a paper with Morrison and Pe-
ters that is accepted in Selecta Mathematica and generalizes a theorem of Kuperberg [Kup94].
Theorem 2. Suppose that a non-degenerate trivalent planar algebra has the dimensions
of the space of n-boundary point diagrams bounded by (1, 0, 1, 1, 4, 10) for n ≤ 5, then it
must be quantum SO(3), quantum G2 , or a certain free product of quantum SU (2) at a
10th root of unity with another quantum SU (2).

Underlying this result, is a simple lemma about planar trivalent graphs (i.e. a bunch of
vertices in the plane connected by edges in a non-intersecting way such that there are ex-
actly three edges coming into each vertex), which states that any planar 3-valent graph
must contain a face that is a pentagon or smaller. This lemma follows from a few lines of
algebra, and an important theorem of Euler’s from the 18th century which says that for
any planar graph the number of vertices minus the number of edges plus the number of
faces is equal to 2 (e.g. for the cube 8 − 12 + 6 = 2). To understand larger trivalent planar
algebras we need generalizations of this lemma. An enormous number of such generaliza-
tions can be proved using a technique called the discharging method.
The discharging method was developed to prove the 4-color theorem, which states that
for any planar graph you can color the vertices using only four colors, such that two ver-
tices connected by an edge always have distinct colors. The four color theorem was fa-
mously proved by Appel–Haken in 1976 using a computer, and was the first major computer-
assisted theorem. They used a very complicated discharging argument, but we are able to
improve our result by using the very simplest application of discharging. Namely, any pla-
nar trivalent graph contains a square or smaller face, or it contains two adjacent pentago-
nal faces, or it contains a pentagonal face adjacent to a hexagonal face. Using this lemma,
together with a new computational software package for skein theory, we proved the fol-
lowing theorem.
Theorem 3. Suppose that a non-degenerate trivalent planar algebra has the dimensions
of the space of n-boundary point diagrams bounded by (1, 0, 1, 1, 4, 11, 40) for n ≤ 6, then
it must be one of the planar algebras from the previous theorem, or the planar algebra H3
closely related to the Haagerup subfactor (which we constructed in [GS12b]).

Our approach to trivalent categories also has interesting applications to the Deligne–
Vogel exceptional series and its quantum analogue, which we are investigating in work in
progress with Morrison and Thurston. Deligne (who won the Fields medal in 1978 and the
Abel prize in 2013) observed that the exceptional Lie groups D4 o Z/3, F4 , E6 o Z/2, E7 ,
and E8 appear to live in a continuous family of symmetric tensor categories [Del96]. Pla-
nar algebraically, Lie algebras can be described by graphs made from trivalent vertices and
symmetric crossings, satisfying certain natural relations. Vogel observed that the following
relation holds for the exceptional series [Vog99].
" #
= + +a + + ,

The following conjecture is one of the most important open questions in planar algebras.
The uniqueness part should be easier than the existence part. We hope to approach it us-
ing discharging combined with other skein theoretic techniques.
Conjecture 4. For generic a, there exists a unique trivalent category satisfying antisym-
metry, the Jacobi relation, and the Vogel relation.

Our main result is formulating the correct quantum version of the above conjecture.
Theorem 5. Any braided tensor category deforming the conjectural exceptional series
must satisfy the following quantum Vogel relation.

" #
v −3 +v − v −1 +α + v −4 + v4 = 0
Conjecture 6. For generic v and α, there exists a unique braided tensor category satisfy-
ing the the quantum Vogel relation and a quantum version of anti-symmetry.

Using the quantum Vogel relation we can compute the quantum analogues of the di-
mension formulas in [CdM96] and the value of the conjectural 2-variable knot polynomial
corresponding to this series for many knots (including all knots with 8 or fewer crossings).
Unconditionally, this gives the best known method for computing the invariants coming
from the large exceptional quantum groups for many links.
Our skein theoretic approach has also suggested two possible candidates for new exotic
tensor categories. The first appears when we considered “twisted” trivalent categories
(where rotating the vertex multiplies by a cube root of unity). The second comes from
looking at the quantum exceptional series specialized to v a 5th root of unity. We plan
to explore these examples in future work. We also plan to use our automated skein theory
software to consider other skein theoretic settings, for example, coming from quadrilaterals
of intermediate subfactors.

4. Tensor categories and local topological field theories

My final research program is quite different in feel from the previous three. A topologi-
cal quantum field theory is a notion from physics which has been used in mathematics to
give connections between algebra (the study of operations like multiplication) and topol-
ogy (the study of spaces like spheres and tori). Recently this field has been revolutionized
by higher category theory, a very abstract branch of mathematics concerning “functions
between functions between functions.” In particular, Lurie proved the Baez–Dolan Cobor-
dism Hypothesis [Lur09, BD95], which gives a higher categorical classification of TQFTs,
for which he received a $3 million Breakthrough Prize. Originally, in work of Reshetikhin,
Turaev, Viro, Witten, and others, TQFTs were used to give applications from quantum
groups to 3-dimensional topology [Tur94, RT91, TV92, Wit89]. More recently, applications
have gone the opposite direction, using topology to better understand algebra. In joint
work with Douglas and Schommer–Pries we use the cobordism hypothesis to give topologi-
cal insight into the structure of tensor categories. This clarifies the appearance of quantum
groups in TQFTs, generalizes Turaev–Viro theories to the framed setting, and gives topo-
logical proofs of several algebraic results.
One striking application of our theory is a topological interpretation of Radford’s theo-
rem for tensor categories, which says that taking two duals need not be trivial, but taking
four duals always is. Using our approach we can deduce Radford’s theorem directly from
a well-known topological fact called “the Dirac belt trick.” The Dirac belt trick is a bit
hard to describe in text, but videos are easy to find on YouTube. Roughly, if you hold
your hand out from your body palm up and then rotate it 360 degrees clockwise while
keeping your palm up, your arm will be twisted, but if you rotate again another 360 de-
grees clockwise then your arm untwists. Under the correspondence between topology and
algebra given by this TQFT each 180 degree rotation corresponds to taking one dual, and
so 720 degree rotation corresponds to taking four duals. This argument, is summarized in
Figure 3, where the “Radford bordism” drawn on the left is input into the topological field
theory attached to a finite tensor category, and the output on the right is a proof of Rad-
ford’s theorem for that tensor category.

= evL

R
= 1evL } v2R

= evL ◦ ev ◦ evR

= v1 } 1evR

= evR

R = (v1 } 1evR ) ◦ (1evL } v2R )

Figure 3. A Morse decomposition of the Radford bordism.

In later work in the series, we will prove several results about the SO(3) action on the
spaces of tensor categories given by changing the framing. In related work in progress with
Ben-Zvi, Brochier and Jordan, we adapt this to the 4-dimensional setting and to the case
of infinite tensor categories. We hope that this construction will give insight into many
open questions in representation theory and topology, for example the AJ conjecture.

5. Relationship to physics and computer science

Over the past 20 years there has been increasing interaction between mathematicians
studying tensor categories, condensed matter physicists studying topological orders of mat-
ter, and computer scientists studying quantum computation. Roughly the idea is that
certain exotic materials called topological orders can be created and studied in the lab
(specifically fractional quantum Hall systems realized in gallium arsenide heterostruc-
tures), that these materials behave in ways that are described by modular tensor cate-
gories, and that one could use these materials to create a quantum computer in a way that
is intrinsically fault-tolerant due to their topological nature. To give some examples of the
level of interest in this area, Laughlin, Störmer, and Tsui received the Nobel Prize in 1998
for their work on the fractional quantum Hall affect, Microsoft Research has a system of
labs called Station Q working on building a topological quantum computer which is run
by Fields medalist Mike Freedman, and Alexei Kitaev’s work on topological quantum com-
putation has been recognized by both a MacArthur fellowship and one of the inaugural
million dollar Fundamental Physics Prizes. This is also an area of active research at Indi-
ana University, connected somewhat to Prof. Ortiz’s Emerging Areas of Research proposal
on Quantum Information Technology. My own work does not mainly focus on this direc-
tion of applications of tensor categories to physics and computer science, but I do stay in-
volved with this field, for example, speaking at an IPAM conference on “Symmetry and
Topology in Quantum Matter” and visiting Station Q several times. We have proposed a
semester program on Quantum Symmetry at MSRI for which we intend to include people
from the quantum information community. Due to engineering difficulties, the examples
that can be currently constructed in the lab appear to relate to the very simplest tensor
categories (corresponding to the A3 and A4 subfactors), but over the next several decades
it is possible that more complicated tensor categories will appear in nature, in which case
my example-driven work may become of broader interest in science.

References
[AH99] Marta Asaeda and√Uffe Haagerup. Exotic subfactors of finite depth with Jones indices (5 +
√
13)/2 and (5 + 17)/2. Comm. Math. Phys., 202(1):1–63, 1999. Preprint arXiv:[Link]/
9803044 MR1686551 DOI:10.1007/s002200050574.
[AMP15] Narjess Afzaly, Scott Morrison, and David Penneys. The classification of subfactors with index
at most 5 14 ., 2015. Preprint arXiv:1509.00038.
[AY09] Marta Asaeda and Seidai Yasuda. On Haagerup’s list of potential principal graphs of sub-
factors. Comm. Math. Phys., 286(3):1141–1157, 2009. Preprint arXiv:0711.4144 MR2472028
DOI:10.1007/s00220-008-0588-0.
[BD95] John C. Baez and James Dolan. Higher-dimensional algebra and topological quantum field
theory. J. Math. Phys., 36(11):6073–6105, 1995. Preprint arXiv:q-alg/9503002 MR1355899
DOI:10.1063/1.531236.
[Bis98] Dietmar Bisch. Principal graphs of subfactors with small Jones index. Math. Ann., 311(2):223–
231, 1998. MR1625762 DOI:10.1007/s002080050185.
[BMPS12] Stephen Bigelow, Scott Morrison, Emily Peters, and Noah Snyder. Constructing the ex-
tended Haagerup planar algebra. Acta Math., 209(1):29–82, 2012. Preprint arXiv:0909.4099
MR2979509 DOI:10.1007/s11511-012-0081-7.
[BP14] Stephen Bigelow and David Penneys. Principal graph stability and the jellyfish algorithm,
2014. Preprint arXiv:1208.1564 MR3157990 DOI:10.1007/s00208-013-0941-2.
[CdM96] Arjeh M. Cohen and Ronald de Man. Computational evidence for Deligne’s conjecture re-
garding exceptional Lie groups. C. R. Acad. Sci. Paris Sér. I Math., 322(5):427–432, 1996.
MR1381778.
[CMS11] Frank Calegari, Scott Morrison, and Noah Snyder. Cyclotomic integers, fusion categories, and
subfactors. Comm. Math. Phys., 303(3):845–896, 2011. Preprint arXiv:1004.0665 MR2786219
DOI:10.1007/s00220-010-1136-2.
[Del96] Pierre Deligne. La série exceptionnelle de groupes de Lie. C. R. Acad. Sci. Paris Sér. I Math.,
322(4):321–326, 1996. MR1378507.
[EG11] David E. Evans and Terry Gannon. The exoticness and realisability of twisted Haagerup-
Izumi modular data. Comm. Math. Phys., 307(2):463–512, 2011. Preprint arXiv:1006.1326
MR2837122 DOI:10.1007/s00220-011-1329-3.
[ENO10] Pavel Etingof, Dmitri Nikshych, and Victor Ostrik. Fusion categories and homotopy theory.
Quantum Topol., 1(3):209–273, 2010. (with an appendix by Ehud Meir), Preprint arXiv:0909.
3140 MR2677836 DOI:10.4171/QT/6.
[GdlHJ89] Frederick M. Goodman, Pierre de la Harpe, and Vaughan F. R. Jones. Coxeter graphs and tow-
ers of algebras, volume 14 of Mathematical Sciences Research Institute Publications. Springer-
Verlag, New York, 1989. MR999799.
[GIS15] Pinhas Grossman, Masaki Izumi, and Noah Snyder. The Asaeda-Haagerup fusion categories,
2015. Preprint arXiv:1501.07324.
[GJS12] Pinhas Grossman, David Jordan, and Noah Snyder. Cyclic extensions of fusion categories via
the brauer-picard groupoid, 2012. Preprint arXiv:1211.6414, to appear in Quantum Topology.
[GS12a] Pinhas Grossman and Noah Snyder. The Brauer-Picard group of the Asaeda-Haagerup fusion
categories, 2012. Preprint arXiv:1202.4396, to appear in Trans. Am. Math. Soc.
[GS12b] Pinhas Grossman and Noah Snyder. Quantum subgroups of the Haagerup fusion cate-
gories. Comm. Math. Phys., 311(3):617–643, 2012. Preprint arXiv:1102.2631 MR2909758
DOI:10.1007/s00220-012-1427-x. √
[Haa94] Uffe Haagerup. Principal graphs of subfactors in the index range 4 < [M : N ] < 3 + 2. In
Subfactors (Kyuzeso, 1993), pages 1–38. World Sci. Publ., River Edge, NJ, 1994. MR1317352.
[Han10] Richard Han. A Construction of the “2221” Planar Algebra. PhD thesis, University of Califor-
nia, Riverside, 2010. Preprint arXiv:1102.2052.
[IJMS12] Masaki Izumi, Vaughan F. R. Jones, Scott Morrison, and Noah Snyder. Subfactors of index
less than 5, Part 3: Quadruple points. Comm. Math. Phys., 316(2):531–554, 2012. Preprint
arXiv:1109.3190 MR2993924 DOI:10.1007/s00220-012-1472-5.
[Izu] Masaki Izumi. Notes on 32n subfactors.
[Izu01] Masaki Izumi. The structure of sectors associated with Longo-Rehren inclusions. II. Examples.
Rev. Math. Phys., 13(5):603–674, 2001. MR1832764 DOI:10.1142/S0129055X01000818.
[Jon83] Vaughan F. R. Jones. Index for subfactors. Invent. Math., 72(1):1–25, 1983. MR696688
DOI:10.1007/BF01389127.
[Jon85] Vaughan F. R. Jones. A polynomial invariant for knots via von Neumann algebras. Bull. Amer.
Math. Soc. (N.S.), 12(1):103–111, 1985. MR766964 DOI:10.1090/S0273-0979-1985-15304-2.
[Kau87] Louis H. Kauffman. State models and the Jones polynomial. Topology, 26(3):395–407, 1987.
MR899057 DOI:10.1016/0040-9383(87)90009-7.
[Kaw95] Yasuyuki Kawahigashi. Classification of paragroup actions in subfactors. Publ. Res. Inst. Math.
Sci., 31(3):481–517, 1995. MR1355948 DOI:10.2977/prims/1195164051.
[Kup94] Greg Kuperberg. The quantum G2 link invariant. Internat. J. Math., 5(1):61–85, 1994.
Preprint arXiv:[Link]/9201302 MR1265145 DOI:10.1142/S0129167X94000048.
[Liu15] Zhengwei Liu. Composed inclusions of A3 and A4 subfactors. Adv. Math., 279:307–371, 2015.
Preprint arXiv:1308.5691 MR3345186 DOI:10.1016/[Link].2015.03.017.
[Lur09] Jacob Lurie. On the classification of topological field theories. In Current developments in
mathematics, 2008, pages 129–280. Int. Press, Somerville, MA, 2009. MR2555928 Preprint
arXiv:0905.0465.
[MPPS12] Scott Morrison, David Penneys, Emily Peters, and Noah Snyder. Classification of subfactors of
index less than 5, part 2: triple points. International Journal of Mathematics, 23(3):1250016,
2012. Preprint arXiv:1007.2240 MR2902285 DOI:10.1142/S0129167X11007586.
[MS12] Scott Morrison and Noah Snyder. Subfactors of index less than 5, Part 1: The principal graph
odometer. Comm. Math. Phys., 312(1):1–35, 2012. Preprint arXiv:1007.1730 MR2914056
DOI:10.1007/s00220-012-1426-y.
[Ocn88] Adrian Ocneanu. Quantized groups, string algebras and Galois theory for algebras. In Operator
algebras and applications, Vol. 2, volume 136 of London Math. Soc. Lecture Note Ser., pages
119–172. Cambridge Univ. Press, Cambridge, 1988. MR996454.
[Ocn01] Adrian Ocneanu. Operator algebras, topology and subgroups of quantum symmetry—
construction of subgroups of quantum groups. In Taniguchi Conference on Mathematics Nara
’98, volume 31 of Adv. Stud. Pure Math., pages 235–263. Math. Soc. Japan, Tokyo, 2001.
MR1865095.
[Pen15] David Penneys. Chirality and principal graph obstructions. Adv. Math., 273:32–55, 2015.
Preprint arXiv:1307.5890 MR3311757 DOI:10.1016/[Link].2014.11.021.
[Pet10] Emily Peters. A planar algebra construction of the Haagerup subfactor. Inter-
nat. J. Math., 21(8):987–1045, 2010. Preprint arXiv:0902.1294 MR2679382
DOI:10.1142/S0129167X10006380.
[Pop94] Sorin Popa. Classification of amenable subfactors of type II. Acta Math., 172(2):163–255, 1994.
MR1278111 DOI:10.1007/BF02392646.
[PT12] David Penneys and James E. Tener. Subfactors of index less than 5, Part 4: Vines.
Internat. J. Math., 23(3):1250017, 18, 2012. Preprint arXiv:1010.3797 MR2902286
DOI:10.1142/S0129167X11007641.
[RT91] Nicolai Reshetikhin and Vladimir G. Turaev. Invariants of 3-manifolds via link polynomials and
quantum groups. Invent. Math., 103(3):547–597, 1991. MR1091619 [Link]/1104180037.
[Tur94] Vladimir G. Turaev. Quantum invariants of knots and 3-manifolds, volume 18 of de Gruyter
Studies in Mathematics. Walter de Gruyter & Co., Berlin, 1994. MR1292673 (preview at google
books).
[TV92] Vladimir G. Turaev and Oleg Ya. Viro. State sum invariants of 3-manifolds and quantum 6j-
symbols. Topology, 31(4):865–902, 1992. MR1191386.
[Vog99] Pierre Vogel. The universal lie algebra, 1999. [Link]
[Wit89] Edward Witten. Quantum field theory and the Jones polynomial. Comm. Math. Phys.,
121(3):351–399, 1989. MR990772 [Link]/1104178138.