The Hochschild cohomology of the group $G^2_3$
Hassan AlHussein, Pavel Kolesnikov

TL;DR
This paper uses discrete algebraic Morse theory to compute the Hochschild cohomology of the group algebra of $G_3^2$, revealing mostly trivial cohomology with specific non-trivial cases.
Contribution
It introduces a novel application of algebraic Morse theory to compute Hochschild cohomology for the group $G_3^2$, providing explicit calculations.
Findings
Most Hochschild cohomology groups are trivial.
Two specific 1-dimensional bimodules have non-trivial $H^2$.
The method simplifies cohomology calculations for this class of groups.
Abstract
We apply discrete algebraic Morse theory to calculate the Anick resolution of the group algebra of the group . As a corollary, we evaluate Hochschild cohomologies of with coefficients in all 1-dimensional bimodules. Almost all these groups are trivial, the only exceptions are 1-dimensional for two particular 1-dimensional bimodules.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAlgebraic structures and combinatorial models · Homotopy and Cohomology in Algebraic Topology · Advanced Operator Algebra Research
The Hochschild cohomology of the group
Hassan AlHussein
Novosibirsk State University, Russia
and
Pavel Kolesnikov
Sobolev Institute of Mathematics, Russia
Abstract.
Cohomology plays an important role in algebra. The group of Hochschild cohomology for group contains information about its structure. In this work we found group of Hochschild -cohomology for the group for any . Anick resolution, Algebraic Morse theory and Gröbner—Shirshov basis were used to get this result.
Key words and phrases:
Hochschild cohomology, Anick resolution, Gröbner–Shirshov basis, Morse matching
1. Introduction
Homological methods allow us to get important information about the structure of an algebra. For associative algebras, Hochschild cohomologies play an important role in structure and representation theory. Finding the Hochschild cohomology group of a given algebra with coefficients in a given -bimodule is often a difficult problem. In order to solve this problem one needs a long exact sequence starting from , a resolution of . The most natural bar-resolution is easy to construct but it is too bulky for computations. Another approach was proposed by David J. Anick in 1986 [1], where it was built a free resolution for associative algebra which is homotopy equivalent to the bar-resolution. The Anick resolution was also used to find Poincare Series [3]. Computation of the differentials in the Anick resolution according to the original algorithm described in [1] is extremely hard. In order to make the computation easier, one may use the discrete algebraic Morse theory based on the concept of a Morse matching defined in [8]. This concept was used in geometry first, then it became applicable in algebra [9, 10].
A series of groups were introduced in [12, 13, 14, 15] in the study of braids, links, and Coxeter groups. In this paper, we consider the group originated in the following (informal) context. Assume three points are moving without collisions on a disk in the plane at time . The trajectories of these three points are characterized by an element of a group constructed as follows. Whenever the points stand in one line, say, the point stands between and (), we write down the generator . Therefore, we have three generators , , and , and the trajectories of three points are characterized by a word in the alphabet . The following relations hold on these generators:
[TABLE]
In the present work, we apply the Morse matching theory to find the Anick resolution and calculate the groups of Hochschild -cohomologies of the group for all with coefficients in all 1-dimensional -bimodules. It is easy to see that all such modules are of the form , where
[TABLE]
for , .
2. Morse matching
In this section, we state main definitions and essential results related to the construction of the Anick chain complex via algebraic Morse theory following [1, 2, 4, 6, 7, 8]. Let be a field and let be a unital associative -algebra with an augmentation, i.e., a -algebra homomorphism . Let be a set of generators for . Suppose that is a well ordering on the free monoid generated by . Denote by the free associative algebra with identity generated by . There is a canonical surjection , so that . Let be a Gröbner—Shirshov basis of , i.e., a confluent set of defining relations for the algebra . Denote by the set of the leading terms of relations from . Following Anick [1], is called the set of obstructions.
For , a word is an -prechain if and only if there exist , , satisfying the following conditions:
- (1)
; 2. (2)
for .
An -prechain is an -chain if only if the integers can be chosen in such a way that is not an -prechain for neither , .
As in [1], we say that the elements of are [math]-chains, the elements of are -chains, and denote the set of -chains by . The cokernel of is denoted by . A word is said to be -reduced if it does not contain a word from as a subword. Since is the set of all obstructions, the set of all non-trivial -reduced words forms a linear basis of [5].
The two-sided bar resolution is a -free bimodule resolution of , where
[TABLE]
An element is denoted by for . The differential is defined as follows:
[TABLE]
Let be a chain complex of free (left) -modules. Choose a basis in each , and set . Then , and the differential may be uniquely presented in the following form:
[TABLE]
where . Construct a directed weighted graph considering as the set of vertices. Edges of are ordered triples of the form , where is the source of , is the range of , and is the weight of . Define the set of weighted edges by the following rule:
[TABLE]
Definition 1**.**
A finite subset is called a Morse matching if and only if
- •
Each vertex lies in at most one edge ;
- •
For all edges , the weight is an invertible element in the center of ;
- •
The graph has no directed cycles, where is given by
[TABLE]
A vertex is critical with respect to if does not belong to an edge ; we denote by the set of critical edges from .
Suppose is a path in a weighted directed graph with vertices . Then
[TABLE]
Denote by , , the sum of weights of all path from to .
Theorem 1**.**
A chain complex is homotopy-equivalent to the complex of free -modules where and
[TABLE]
where is calculated in the graph .
Let be the bimodule bar complex for . We may consider as a complex of left -modules and construct a Morse matching in in the following way.
Theorem 2**.**
For let , be the set of vertices in such that , all are -reduced, and is the largest integer such that . Let ; define a partial matching by letting consist of all edges
[TABLE]
for , where and . Then the set of edges is a Morse matching on with critical cells for all .
Proposition 1**.**
The chain complex is the -free Anick resolution.
Definition 2**.**
Let be a unital -algebra, and let be a -bimodule. Consider the cochain complex
[TABLE]
where
[TABLE]
.
The th Hochschild cohomology group of with coefficients in is .
We may use the Anick resolution to calculate by means of the following diagram:
[TABLE]
Here
[TABLE]
so .
3. The Anick complex for
Definition 3** ([12]).**
The group is defined by generators , and relations , i.e., , where
[TABLE]
Obviously, the set defines the same ideal in as .
Theorem 3** ([11]).**
The set of relations is a confluent set of rewriting rules.
To be more precise, Theorem 3 states that is a Gröbner—Shirshov basis of the group algebra with respect to the tower order on , where .
To find Anick complex for , we need several steps. First, we have to find the set of obstructions for the group relative to the given Gröbner—Shirshov basis (the set of leading terms in ) and the set of -chains for . Next, we should build a Morse matching in the graph and construct the graph . Finally, it remains to calculate the Anick differential.
In order to make notations shorter, we will often use for an element , where is an ()-chain.
Example 1**.**
For , . Let us construct Morse matching and evaluate as described in Theorem 2. The simplest case is
[TABLE]
Hence, . Similarly, , . A more complicated construction is needed for :
* [ca]\stackrel{{\scriptstyle b\otimes 1}}{{\longleftarrow}}[b|ca]\stackrel{{\scriptstyle 1\otimes ca}}{{\longrightarrow}}[b]$$-1$$1$$[c|a]$$1\otimes a$$[c]$$c\otimes 1$$[a]$$-1$$[acb]$$-1$$1$$[cb]\stackrel{{\scriptstyle a\otimes 1}}{{\longleftarrow}}[a|cb]\stackrel{{\scriptstyle 1\otimes cb}}{{\longrightarrow}}[a]$$-1$$1$$[b]\stackrel{{\scriptstyle c\otimes 1}}{{\longleftarrow}}[c|b]\stackrel{{\scriptstyle 1\otimes b}}{{\longrightarrow}}[c] *
Hence, . Similarly, .
Example 2**.**
For , . Let us construct Morse matching and evaluate . The simplest case is
[TABLE]
Hence, . Similarly, , . A more complicated construction is needed for :
* [ca|a]\stackrel{{\scriptstyle b\otimes 1}}{{\longleftarrow}}[b|ca|a]\stackrel{{\scriptstyle-1\otimes a}}{{\longrightarrow}}[b|ca]$$-1$$1$$[a|a]\stackrel{{\scriptstyle c\otimes 1}}{{\longleftarrow}}[c|a|a]$$-1$$[acb|a]$$-1$$1$$[cb|a]\stackrel{{\scriptstyle a\otimes 1}}{{\longleftarrow}}[a|cb|a]$$1$$[a|abc]$$-1$$1$$[a|a|bc]\stackrel{{\scriptstyle-1\otimes bc}}{{\longrightarrow}}[a|a]$$1$$-1$$[b|a]\stackrel{{\scriptstyle c\otimes 1}}{{\longleftarrow}}[c|b|a]$$1$$[c|cabc]$$1$$-1$$[c|c|abc]$$-1\otimes abc$$[c|c] *
Hence, . Similarly, , , .
Theorem 4**.**
For , . Then the Anick differential is given by the following rules.
[TABLE]
When is an even number,
[TABLE]
if is even then
[TABLE]
if is odd then
[TABLE]
When is an odd number,
[TABLE]
if is even then
[TABLE]
if is odd then
[TABLE]
Proof. It is enough to construct a Morse matching for the graph corresponding to the bar resolution and calculate the differentials in the graph . As in Examples 1, 2, this can be done by the algorithm described in Theorem 2. Namely, for each we need to construct the subgraph of which contains all paths from to vertices , , and calculate the differentials (2).
Figure 1 presents the subgraph which allows us to calculate . For convenience, vertices from are boxed. For , , is even, is odd, the corresponding graph is shown on Figures 2–4. Other elements of can be processed in a similar way.∎
4. Hochschild cohomologies of
Suppose is a 1-dimensional bimodule over . Then , and there exist such that , for , . Denote the bimodule obtained by .
It is natural to identify with a triple of signs, e.g., if , , , then
[TABLE]
Our aim is to calculate for all , . Up to the natural equivalence, it is enough to consider the following bimodules:
- (1)
; 2. (2)
; 3. (3)
; 4. (4)
; 5. (5)
; 6. (6)
; 7. (7)
; 8. (8)
.
For every above, the map from Theorem 4 (for , see Examples 1, 2) induces a linear map from to , let us denote it by . The corresponding conjugate map acts from the space to , and
[TABLE]
so
[TABLE]
Since , it is enough to evaluate which is straightforward.
The group is generated by involutions, so the first Hochschild cohomology group is expectedly trivial.
Corollary 1**.**
For , we have , but for the second Hochschild cohomology group is nontrivial, .
Corollary 2**.**
For , for all 1-dimensional -bimodules.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] D. J. Anick, On the homology of associative algebras, Transactions of the American Mathematical Society, 1986, Vol. 296, No 2, P. 641-659.
- 2[2] V. Lopatkin, Cohomology rings of the plactic monoid algebra via a Gröbner-Shirshov basis, J. Algebra Appl. 15 (5) (2016), 1650082, 30 pp.
- 3[3] V. A. Ufnarovsky, Combinatorial and asymptotic methods in algebra. (Russian) Current problems in mathematics. Fundamental directions, Vol. 57 (Russian), 5–177, Moscow, 1990.
- 4[4] S. Maclane, Homology, Springer Verl., Berlin–Gottingen–Heideiberg, 1963.
- 5[5] L. A. Bokut. Imbeddings into simple associative algebras (Russian), Algebra i Logika 15 (1976) 117–142.
- 6[6] L. A. Bokut, Y. Chen, Gröbner-Shirshov bases and their calculation, Bull. Math. Sci. 4 (3) (2014) 325–395.
- 7[7] E. Sköldberg, Morse theory from an algebraic viewpoint, Trans. Amer. Math. Soc., 358 (1) (2006) 115–129.
- 8[8] M. Jöllenbeck, V. Welker, Minimal Resolutions Via Algebraic Discrete Morse Theory, Mem. Am. Math. Soc. 197 (2009), no. 923.
