Introduction
In the standard formulation of quantum mechanics, the Schrödinger picture describes the evolution of a single global wavefunction. While this approach is mathematically elegant, it often obscures spatial locality—the principle that physical processes in one region of space should not instantaneously affect distant regions. In contrast, the Heisenberg picture makes locality explicit by evolving local operators, but it lacks the intuitive appeal of a state vector representation.
The "gauge picture" was recently proposed as a middle ground. It maintains the use of wavefunctions but localizes them to specific spatial patches. However, a significant drawback of the original gauge picture is that these local states tend to "scramble" over time, deviating significantly from the global Schrödinger state. This work introduces a tunable interpolation method to bridge these two representations, allowing local wavefunctions to remain aligned with the global state while preserving explicit locality in the equations of motion.
The Gauge Picture Framework
The gauge picture decomposes a quantum system into a collection of overlapping spatial patches. Each patch $I$ is assigned its own local wavefunction $|\psi_I\rangle$. To maintain consistency between these patches, time-dependent unitary connections $U_{IJ}$ are introduced to relate the Hilbert spaces of adjacent regions.
While this framework ensures that the equations of motion are spatially local, the local wavefunctions $|\psi_I\rangle$ do not naturally correspond to the global Schrödinger wavefunction $|\Psi\rangle$. In the absence of additional constraints, the connections $U_{IJ}$ evolve away from the identity, and the local states become increasingly "scrambled" relative to the global state.
Methods: The Interpolation Term
To address this, the authors modify the gauge picture's equations of motion by adding a new term, $\gamma X_I$, to the generator of time evolution. Here, $\gamma$ is a real-valued constant that acts as a "dial" for interpolation.
The operator $X_I$ is derived using partial traces and is designed to penalize the connections $U_{IJ}$ for deviating from the identity matrix. When $\gamma$ is large, this term effectively "pulls" the local gauge-picture wavefunctions toward the global Schrödinger-picture wavefunction. In the limit where $\gamma \to \infty$, the connections approach the identity, and all local wavefunctions converge to the global state.
The researchers tested this interpolation using numerical simulations of the 1D Transverse-Field Ising Model (TFIM), a standard benchmark in many-body physics. They explored both integrable (solvable) and non-integrable (chaotic) regimes by varying the longitudinal field $h_z$ and simulating chain lengths from $L=5$ to $L=10$.
Key Results and Scaling Analysis
The study reveals a fundamental trade-off between locality and global state alignment. The most critical finding is the "price of locality": to keep the local wavefunctions matching the global one as the system size $L$ increases, the coefficient $\gamma$ must scale exponentially ($e^{aL}$).
Numerical data confirmed that the deviation $S_{IJ}$ between the gauge and Schrödinger pictures follows a specific scaling law: $S_{IJ} \sim \gamma^{-2} e^{aL+b}$. This $1/\gamma^2$ dependence matches the authors' perturbative theoretical predictions and highlights the difficulty of maintaining a global state representation in a local framework for large systems.
Furthermore, the dynamics exhibit a sharp transition. At low values of $\gamma$, the deviation between the pictures oscillates erratically—a phenomenon the authors call "squiggles." As $\gamma$ increases beyond a critical threshold $\gamma_0$, these oscillations vanish, and the system enters a stable steady state. The time $t_s$ at which these squiggles begin was found to diverge as $\gamma$ approaches $\gamma_0$, indicating a phase-like transition in the representation's stability.
Non-linear Chaos and Numerical Stability
Unlike the linear Schrödinger equation, the gauge picture equations of motion are inherently non-linear. This non-linearity introduces a "butterfly effect" where tiny numerical errors in integration grow exponentially over time. In simulations, this led to a divergence from the exact mathematical solution after a characteristic time (approximately $t \approx 18$ in the tested models). This "gauge picture chaos" highlights a unique challenge in using local wavefunction representations for long-time simulations, as the non-linearity amplifies integration errors that would remain bounded in the standard Schrödinger picture.
Conclusion
This work demonstrates that it is possible to interpolate between the gauge and Schrödinger pictures, providing a way to maintain "Schrödinger-like" states within a locally explicit framework. However, the exponential scaling of the required interpolation strength suggests a significant complexity barrier for large-scale systems. These findings open new avenues for investigating the stability of different mathematical representations of quantum dynamics and the fundamental costs associated with enforcing spatial locality in state-vector descriptions.
Summary updated