arXiv:2610.01782
Non-Markovian effects on informational steady states
Jacob Werner
Abstract
Informational steady states (ISSs) arise when the rate at which information is gained from continuous measurements is exactly balanced by the rate at which information from earlier measurements is lost to the environment. While ISSs have been studied in Markovian settings, relatively little is known about how memory effects influence their formation and steady-state properties. In this work, we investigate non-Markovian effects on ISSs within the framework of continuously monitored collision models. We model the environment as a sequence of two-qubit ancillas and find that, for this specific model, the steady-state information gain per measurement is negatively correlated with the degree of non-Markovianity. Numerical results further characterize how the system-ancilla and ancilla-ancilla interaction parameters influence both the degree of non-Markovianity and the steady-state information gain.
AI Summary
The paper asks how environmental memory affects informational steady states (ISSs), where continuous monitoring produces nonzero information gain balanced by loss of information from earlier measurements. It studies a qubit collision model with two-qubit ancillas, initialized in a diagonal thermal state with population $0.3$ and in $|+\rangle$, respectively. Partial-swap system–ancilla and ancilla–ancilla interactions control the dynamics; projective measurements on the first ancilla qubit generate the measurement record. The analysis decomposes the change in system–record Holevo information as $\Delta I_t=G_t-L_t$ and numerically compares the steady-state measurement gain $G_\infty$ with trace-distance backflow. Parameter scans yield negative Spearman correlations of approximately $-0.33$, $-0.52$, and $-0.31$, with reported $p<10^{-20}$. Both ancilla–ancilla couplings generally need to be strong for substantial backflow. Gain varies nonmonotonically: correlations transferred between ancillas can partly restore system–measured-ancilla correlations during subsequent interactions. Near full ancilla swaps, backflow is largest when the two system–ancilla swap angles sum to approximately $\pi/2$; an ideal limiting case gives a divergent cumulative backflow measure. These results concern one interaction architecture and fixed ancilla preparations and measurements. Gain estimates average 20 trajectories of 100 collisions over their final 30 steps, reflecting stated computational constraints. The BLP maximization is replaced by a fixed pair of $\sigma_x$ eigenstates. The paper emphasizes that backflow characterizes the transient dynamics whereas gain is evaluated at steady state, and that their correlation is not causal or necessarily general: other models could exhibit positive correlations.
Central claim may not hold: The reported correlations use a fixed-pair trace-distance backflow witness rather than a verified optimized BLP measure, leaving their interpretation as correlations with BLP non-Markovianity unestablished. The identification of a finite-window, 20-trajectory gain estimate with the asymptotic steady-state gain lacks convergence checks and uncertainty estimates, especially near strongly memory-preserving parameters. The argument that bounded Holevo information and time-independent dynamics guarantee vanishing information increments is invalid, and the paper itself describes a periodic system-dynamics limit.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.