In "On the Classification of Topological Field Theories" in Example 1.4.1, Lurie introduces the B-model with target an (even dimensional) Calabi-Yau variety $X$: The Hochschild cohomology $\operatorname{HH}^*(X)$, together with the "canonical trace map" $\operatorname{HH}^*(X) \rightarrow k$, form a graded commutative Frobenius algebra; and by a classical theorem that can, as he states, analogously applied to the graded setting, this defines a 2-dimensional Topological Quantum Field Theory $Z$ with values in complexes such that $Z(S^1)=\operatorname{HH}^*(X)$.

While I can roughly follow his reasoning (although it would be helpful if someone could provide a short explanation on how this trace map occurs), as far as I knew the states of the B-model on $X$ were related to the (bounded) derived category $\operatorname{D}^b(\operatorname{Coh(X)})$, and I don't really see why Lurie's definition should describe the B-model as I know it, or why Hochschild cohomology occurs here. As Lurie only uses this statement as a motivation for introducing extended TQFTs with values in chain complexes, he doesn't give any further explanation or references, so I wondered if anyone here could. Thanks in advance and greetings,

Markus Zetto