Trainability · Primitives

g-sim: transfer-matrix coordinate propagation

Theorem statement

With respect to a Hilbert–Schmidt-orthonormal Hermitian basis $\{B_j\}_{j=1}^{\dim\mathfrak g}$ of $i\mathfrak g$, write the coordinate vector $c(X)=\big(\langle B_j,X\rangle_{\mathrm{HS}}\big)_{j=1}^{\dim\mathfrak g}$. For a circuit $U=\prod_{k=1}^{L}e^{i A_k}$ with generators in $\mathfrak g$, the coordinates of the Heisenberg-evolved observable follow from those of $O$ by an ordered product of per-gate $\dim\mathfrak g\times\dim\mathfrak g$ transfer matrices, $$c\big(\mathcal H(O)\big)=G_L\cdots G_1\,c(O),\qquad (G_k)_{ij}=\big\langle B_i,\,e^{-i A_k}B_j\,e^{i A_k}\big\rangle_{\mathrm{HS}}=\big(e^{-i\,\operatorname{ad}_{A_k}}\big)_{ij},$$ where $\operatorname{ad}_{A_k}(X)=[A_k,X]$. Each $G_k$ is the adjoint action of one gate written in the basis, so the entire g-sim update runs in the $\dim\mathfrak g$-dimensional coordinate space.

Sources

  1. Lie-algebraic classical simulations for quantum computing

    Matthew L. Goh, Martin Larocca, Lukasz Cincio, M. Cerezo, Frederic Sauvage, 2023

  2. Does provable absence of barren plateaus imply classical simulability? Or, why we need to rethink variational quantum computing

    M. Cerezo, Martin Larocca, Diego Garcia-Martin, N. L. Diaz, Paolo Braccia, Enrico Fontana, Manuel S. Rudolph, Pablo Bermejo, Aroosa Ijaz, Supanut Thanasilp, Eric R. Anschuetz, Zoe Holmes, 2023

Lean context

Copy a short prompt with the import, theorem name, citations, and public source link.

Open Lean source