Higher Hopf formula for homology of Leibniz \(n\)-algebras (Q847678)

There is a well known correspondence in homotopical algebra between four types of objects. Let my state it here in terms of groups. Namely, crossed modules of groups correspond to \(1\)-cat groups, correspond to group objects in the category of categories and to simplicial groups with Moore complex of length one. One early reference for this is \textit{J.-L. Loday}'s paper ``Spaces with finitely many non-trivial homotopy groups'' [J. Pure Appl. Algebra 24, 179--202 (1982; Zbl 0491.55004)]. This correspondence goes as follows: given a crossed module of groups \(\mu:M\to N\), one defines a small category in the category of groups \({\mathtt Grp}\) by taking as set of objects the group \(N\) and set of morphisms the semi-direct product group \(M\rtimes N\), \(N\) acting on \(M\) by the given action in the crossed module. Source and target projections are given by the group morphisms \(s(m,n)=n\) and \(t(m,n)=\mu(m)n\). On the other hand, the two commuting subgroups in the \(1\)-cat group \(G:=M\ltimes N\) are \(\text{ker}(s)\) and \(\text{ker}(t)\). The commutation comes from the double four exchange property of a category object in \({\mathtt Grp}\). On the third side, the nerve of a small category being a simplicial set, we get a simplicial group \(K_*\) out of a small category in \({\mathtt Grp}\). Its Moore complex is then by definition the restriction of the differential \(d_0\) to the intersection \(\bigcap_{i=1}^n\text{ker}(d_i)\) of the kernels of the remaining differentials in \(K_n\). In our case, it is clearly of length one. In the same way, techniques in homotopical algebra have been developed to express \(n\)-cat groups. The objects replacing crossed modules are then called crossed \(n\)-fold extensions or crossed \(n\)-cubes, see \textit{J. Huebschmann} [``Crossed \(n\)-fold extensions of groups and cohomology'', Comment. Math. Helv. 55, 302--313 (1980; Zbl 0443.18019)]. Similarly to what has been said before, the simplicial objects arising here are constructed from iterated semi-direct products \(M\rtimes\ldots\rtimes N\), projecting down to the different factors like in descent theory. It is therefore natural to understand crossed \(n\)-cubes in terms of Čech resolutions. In the article under review, these techniques are applied to Leibniz \(n\)-algebras, i.e., to vector spaces \({\mathcal L}\) with \(n\)-ary brackets \([-,\ldots,-]:{\mathcal L}^{\otimes n}\to{\mathcal L}\) satisfying the identity \[ [[x_1,\dots,x_n],y_1,\dots,y_{n-1}]\,=\,\sum_{i=1}^n[x_1,\dots,x_{i-1}, [x_i,y_1\dots,y_{n-1}],x_{i+1},\ldots,x_n]. \] For \(n=2\), one gets back Leibniz algebras, the non-necessarily skewsymmetric version of Lie algebras. The key to the homological properties of Leibniz \(n\)-algebras is the fact that for a given Leibniz \(n\)-algebra \({\mathcal L}\), there is a Leibniz algebra structure on \({\mathcal D}_{n-1}({\mathcal L}):={\mathcal L}^{\otimes n-1}\) such that \({\mathcal L}\) is a Leibniz \({\mathcal D}_{n-1}({\mathcal L})\)-module and \[ HL_0({\mathcal D}_{n-1}({\mathcal L}),{\mathcal L})\,\cong\,\,{\mathfrak A} {\mathfrak b}({\mathcal L}), \] where \({\mathfrak A}{\mathfrak b}({\mathcal L})\) is the abelianization of \({\mathcal L}\) and the LHS means the Leibniz homology of \({\mathcal D}_{n-1}({\mathcal L})\) with values in \({\mathcal L}\). Indeed, the above fact implies that the Leibniz \(n\)-algebra homology (defined as \(HL_*({\mathcal D}_{n-1}({\mathcal L}),{\mathcal L})\)) is the derived functor of the abelianization (Theorem \(4\)), i.e., is Quillen homology. Next comes the definition of crossed \(m\)-cubes of Leibniz \(n\)-algebras in Section \(4\) leading to an equivalence between the category of crossed \(m\)-cubes of Leibniz \(n\)-algebras and the category of \(m\)-cat Leibniz \(n\)-algebras (Theorem \(8\)). The associated \(m\)-simplicial objects and their diagonals finally lead to the higher Hopf formula for the homology of Leibniz \(n\)-algebras (Theorem \(17\)), expressing the homology groups in terms of the combinatorics of a free simplicial resolution.