arXiv:2610.04075
Gate-Based Quantum Simulation of Finite-Temperature Open-System Dynamics: A Post-Processing Reconstruction Approach for Collision Models
Kazunari Hashimoto, Chikako Uchiyama
Abstract
Simulating open-system dynamics on gate-based quantum computers offers a route to exploring dissipative quantum phenomena through programmable quantum processing. Collision models provide a direct and flexible framework by representing the environment through sequential system--ancilla collisions. At finite temperatures, however, they typically require repeated preparation of thermal ancillas, adding circuit overhead and gate errors. Here, we show that, for a broad class of $U(1)$-symmetric system--ancilla interactions, finite-temperature dynamics can instead be reconstructed from simulations with pure ancillas in orthogonal basis states. The target temperature enters only through classical post-processing of the resulting pure-branch data, allowing the same data to be reused for different temperatures without additional quantum processing. Bypassing explicit thermal-state preparation reduces ancilla-preparation overhead and preparation-induced errors.
AI Summary
The paper asks whether finite-temperature collision-model dynamics can be reconstructed from temperature-independent quantum-circuit data, avoiding repeated preparation of thermal ancillas. It considers a system qubit undergoing identical collisions with fresh ancilla qubits through an excitation-conserving interaction, $[H,\sigma_z^S+\sigma_z^A]=0$. By parametrizing the general symmetry-compatible two-qubit unitary, it establishes that the pure-ancilla branches have a common population-relaxation coefficient $0\leq\gamma\leq1$ and coherence multipliers $\lambda_s=\sqrt{\gamma}e^{i\theta_s}$. After any number $n$ of collisions, thermal populations obey the exact reconstruction $z_n^{\rm th}=p_\uparrow z_n^{(\uparrow)}+p_\downarrow z_n^{(\downarrow)}$. Thermal coherence instead requires a correction to either single branch: $w_n^{\rm th}=q^ne^{in(\Theta-\theta_s)}w_n^{(s)}$, where $qe^{i\Theta}=p_\uparrow e^{i\theta_\uparrow}+p_\downarrow e^{i\theta_\downarrow}$ and $0\leq q\leq1$. Thus the same branch data can be reused across temperatures. Compared with purification requiring a rotation, entangling gate, and auxiliary qubit, branch preparation requires at most an $X$ gate. Under stationary, weak, diagonal-preserving preparation noise, population errors scale as $O(p_R+p_{CX})$ versus $O(p_X)$; relative coherence errors scale as $O[n(p_R+p_{CX})]$ versus $O(np_X)$, with no additional preparation-gate contribution when using the hardware-default branch. These results follow analytically, with depolarizing and amplitude-damping examples. The treatment is restricted to memoryless qubit–qubit collisions; non-Markovian extensions remain open. Collision-gate, reset, readout, sampling, and parameter-estimation errors are not separately analyzed. Population reconstruction can extend beyond the stated symmetry when the branches retain a common affine linear part.
Error found: Appendix B.2 incorrectly makes the down-branch coherence eigenvalue depend on beta rather than alpha, producing incorrect reconstruction even for a displayed parameter choice where the collision unitary is the identity. Appendix B.2 incorrectly equates the channel eigenvalue on sigma_+ with the multiplier for w = x + iy, although these are complex conjugates, giving incorrect analytical phase corrections.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.