arXiv:2610.06541
Products of unitaries as continuous evolutions at finite precision
Michael Jarret
Abstract
I present a short approach to discretizing continuous unitary evolution. It gives a correspondence in both directions between continuous unitary evolutions and products of unitaries in finite dimensions, and I use it to reduce discrete adiabatic theorems to continuous ones. At fine steps, holding truncated samples of a regulated Hamiltonian on a suitable finite partition gives an approximating product. Conversely, rounding the clock identifies a product of $T$ unit-step samples with a continuous evolution, up to two half-step phases. I bound its operator-norm distance from the original evolution by $O(1/T)$ through cancellation within each time step. This assumes bounded first and second derivatives, bounded spectral width, and eigenvalue differences bounded away from nonzero multiples of $2π$, uniformly in $T$. For a unitary walk with these derivative bounds, a fixed common free spectral arc supplies a Hermitian logarithm path satisfying the comparison hypotheses. Continuous simulation guarantees then transfer to the phase-corrected product, and continuous adiabatic bounds to the product itself, with an explicit additive $O(1/T)$ error and each adiabatic theorem's hypotheses retained. For positive definite quantum linear systems of condition number $κ$, this comparison and continuous adiabatic theory give a rescaled qubitized walk with $O(κ/δ)$ steps for leakage $δ$, matching the scaling previously obtained through a discrete adiabatic theorem. These products need no discrete adiabatic theorem.
AI Summary
The paper asks when sampled products of unitaries can be compared with continuous evolution closely enough to transfer continuous adiabatic bounds. It treats both fine-step discretization and products whose factors represent unit-time steps. For a regulated, $k$-sparse Hamiltonian on $[0,T]$, piecewise-constant entrywise $b$-bit truncation gives operator-norm evolution error at most $\sqrt{2}k2^{-b}T$. Holding truncated midpoint samples between equal-variation checkpoints instead gives error $\varepsilon$ using $O(TV/\varepsilon)$ factors, where $V$ is operator-norm total variation. These estimates use Duhamel’s formula; Davis–Kahan bounds account for replacing endpoint spectral projectors by truncated ones. The main converse result compares $W_T=e^{-iH((T-1)/T)}\cdots e^{-iH(0)}$ with evolution $U(T)$ generated by $H(t/T)$. After endpoint half-step phases, their distance is $O((L+L^2+M)/T)$, with constants depending on spectral width and separation from nonzero $2\pi$ resonances; $L$ and $M$ bound first and second derivatives. A cellwise correction cancels linear drift and nearly cancels between neighboring cells. Continuous adiabatic leakage bounds therefore transfer with additive $O(1/T)$ error. A common spectral-free arc supplies a logarithm path for unitary walks. Applications recover $O(\kappa/\delta)$ steps or controlled block-encoding queries for positive-definite quantum linear systems, matching previously obtained scaling. The comparison requires uniform regularity and nonresonance conditions. It does not preserve faster-than-$1/T$ continuous leakage decay. Checkpoint computation, factor implementation, and endpoint state preparation or testing entail costs outside the comparison bounds.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 26 pages, 0 figures