arXiv:2609.29385
Anisotropic information geometry of entropy production
Tomohiro Nishiyama, Yoshihiko Hasegawa
Abstract
We reveal that the information geometry of entropy production exhibits an intrinsic anisotropy, providing a unified origin for the inequivalent geometric constraints obeyed by its different components. Using a hierarchical projection structure of quantum relative entropy, we derive an exact orthogonal decomposition of entropy production into three distinct geometric contributions associated with the deviation of the Gibbs projection temperature from the reference temperature, the deviation of the environment from its Gibbs projection state, and system--environment correlations. We demonstrate that these contributions obey fundamentally different geometric bounds: the first reduces to a classical Kullback--Leibler divergence and admits no universal upper bound in terms of the trace distance and the system and environment dimensions alone, whereas the latter two admit rigorous distance-based bounds with dimension-dependent factors. The framework also applies to classical Markovian dynamics described solely in terms of the system state, for which the system--environment correlation contribution is absent from the decomposition. Combining this geometric decomposition with dynamical speed limits, we obtain rigorous bounds on entropy production expressed solely in terms of physically accessible quantities, such as the mean energy and Hamiltonian variance.
AI Summary
The paper asks how different contributions to entropy production are constrained by state-space distance. It considers finite-dimensional system–environment dynamics under joint unitary evolution, allowing initially correlated and nonthermal states and a fixed reference inverse temperature $\beta$. An energy-matched Gibbs state $\gamma_E(\beta^*)$ defines a projection of the environment state onto the Gibbs family. Combining the relative-entropy Pythagorean identity with the mutual-information decomposition gives $D(\rho_{SE}\Vert\rho_S\otimes\gamma_E(\beta))=D_X+D_Y+D_Z$: temperature mismatch, environmental deviation from its Gibbs projection, and system–environment correlations. Entropy production is the change in this sum, rather than necessarily a sum of nonnegative changes. All three components satisfy Pinsker lower bounds $D_i\geq2\mathcal T_i^2$. The temperature component is a classical KL divergence with no universal finite upper bound depending only on trace distance and dimensions. By contrast, $D_Y\leq\mathcal T_Y\ln(d_E-1)+h_b(\mathcal T_Y)$ and $D_Z\leq2\mathcal T_Z\min(\ln d_S,\ln d_E)+g(\mathcal T_Z)$, where $g(x)=(1+x)\ln(1+x)-x\ln x$. These follow from entropy-continuity inequalities. Combining continuity bounds with quantum speed limits bounds $|\Sigma-\beta\Delta\langle H_E\rangle|$ using integrated Hamiltonian variance. A classical single-bath, detailed-balance Markov-jump formulation gives analogous bounds using dynamical activity. Multiple-reservoir extensions and a finite fermionic resonant-level example are included. The quantum setting assumes a time-independent environment Hamiltonian; the stated sine-form speed bounds apply only within specified angle ranges. The numerical example illustrates the decomposition, not bound saturation.
Error found: The classical speed-limit bound is undefined as printed because its activity is a positive constant under the supplied definition, making the integral of sqrt(A(t))/t diverge, and its domain condition uses a different integrand. The asserted existence of a finite Gibbs projection temperature for arbitrary states excludes states at the minimum or maximum mean energy, which require an explicitly allowed infinite-temperature-parameter limit.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 11 pages, 3 figures