The Hochschild cohomology of a Poincare algebra

In this note, we define the notion of a cactus set, and show that its geometric realization is naturally an algebra over Voronov’s cactus operad, which is equivalent to the framed 2-dimensional little disks operad $\mathcal{D}_2$. Using this, we show that the Hochschild cohomology of a Poincar'e algebra A is an algebra over (the chain complexes of) $\mathcal{D}_2$.

💡 Research Summary

The paper introduces a new combinatorial-topological object called a “cactus set” and shows that its geometric realization carries a natural action of Voronov’s cactus operad 𝒞. A cactus set is a refinement of the usual simplicial set: for each degree n one considers a configuration of n discs that intersect in a tree‑like “cactus” pattern. The face maps correspond to merging two adjacent “branches” of the cactus, while the degeneracy maps correspond to splitting a branch. By checking that these operations satisfy the operadic composition rules of 𝒞, the author proves that |X|, the realization of any cactus set X, is automatically a 𝒞‑algebra.

The second major step is to recall that Voronov’s cactus operad is equivalent (as an operad in topological spaces) to the framed little 2‑disk operad 𝔇₂. The operad 𝔇₂ consists of configurations of finitely many disjoint little disks inside a unit disk, each equipped with a rotation angle (the “frame”). The equivalence is given by a concrete map: each branch of a cactus is sent to a little disk, and the angular position of the branch becomes the framing of that disk. This map respects operadic composition, establishing a homotopy equivalence 𝒞 ≃ 𝔇₂.

With this equivalence in hand, the author turns to a Poincaré algebra A. By definition, a Poincar‑é algebra is a finite‑dimensional, graded, commutative algebra equipped with a non‑degenerate, graded‑symmetric bilinear form that identifies the algebra with its linear dual. Classical examples include the cohomology ring of a closed oriented manifold. The Hochschild cochain complex C⁎(A,A) of A carries the usual cup product and Gerstenhaber bracket, making its cohomology HH⁎(A) a Gerstenhaber algebra. Moreover, the Poincaré duality provides a BV (Batalin–Vilkovisky) operator Δ on HH⁎(A).

The core construction lifts the Hochschild cochain complex to a cactus set. Concretely, an n‑cochain (a multilinear map A^{⊗n}→A) is assigned to a cactus with n branches; the operadic face maps act by inserting the multiplication of A along a branch, which coincides with the Hochschild differential. This endows C⁎(A,A) with a 𝒞‑algebra structure. Via the operadic equivalence 𝒞 ≃ 𝔇₂, the same structure yields an action of the framed little 2‑disk operad on the chain level.

The main theorem, therefore, states:

Theorem. If A is a Poincaré algebra, then its Hochschild cohomology HH⁎(A) carries a natural algebra structure over the chain operad of the framed little 2‑disk operad 𝔇₂.

The proof proceeds by constructing explicit chain‑level maps that intertwine the Hochschild differential with the operadic composition in 𝒞, then transporting them across the equivalence to 𝔇₂. In particular, the cup product corresponds to the operadic 1‑ary operation (inserting one disk into another), the Gerstenhaber bracket to the 2‑ary operation, and the BV operator Δ to the rotation of a disk (the framing). The author verifies all operadic relations hold up to homotopy, ensuring that HH⁎(A) is indeed a homotopy 𝔇₂‑algebra.

Finally, the paper discusses implications. Since 𝔇₂‑algebras are precisely the algebraic structures appearing in 2‑dimensional topological field theories and string topology, the result bridges Hochschild cohomology of Poincaré algebras with these geometric contexts. It refines the classical Gerstenhaber‑BV picture by embedding it into the richer operadic framework of framed little disks, opening avenues for further exploration of higher‑genus operations, factorization homology, and connections to quantum field theory.