Abstract:Observational entropy - a quantity that unifies Boltzmann's entropy, Gibbs' entropy, von Neumann's macroscopic entropy, and the diagonal entropy - has recently been argued to play a key role in a modern formulation of statistical mechanics. Here, relying on algebraic techniques taken from Petz's theory of statistical sufficiency and on a Levy-type concentration bound, we prove rigorous theorems showing how the observational entropy of a system undergoing a unitary evolution chosen at random tends to increase with overwhelming probability and to reach its maximum very quickly. More precisely, we show that for any observation that is sufficiently coarse with respect to the size of the system, regardless of the initial state of the system (be it pure or mixed), random evolution renders its state practically indistinguishable from the microcanonical distribution with a probability approaching one as the size of the system grows. The same conclusion holds not only for random evolutions sampled according to the unitarily invariant Haar distribution, but also for approximate 2-designs, which are thought to provide a more physically reasonable way to model random evolutions.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is about the behavior of observational entropy (OE) in isolated systems, especially its increasing trend under random unitary evolution and the probability of reaching its maximum value. Specifically, through rigorous mathematical proofs, the paper shows that in an isolated system, when the system undergoes a randomly - selected unitary evolution, the observational entropy tends to increase with overwhelming probability and quickly reaches its maximum value. This conclusion is applicable not only to random evolutions sampled from the Haar distribution but also to approximate 2 - designs, which are considered a more physically reasonable model of random evolution.
### Main research contents:
1. **Increase in observational entropy in macroscopic states**:
- The paper first considers the case where the system is initially in a macroscopic state. In this case, if the initial state of the system is a macroscopic state, only unitary operators belonging to the zero - volume sub - cluster can keep the observational entropy unchanged; otherwise, the observational entropy will strictly increase. This part of the proof is based on Petz's theory of statistical sufficiency.
2. **Increase in observational entropy for any initial state**:
- Next, the paper uses the Lévy - type concentration bound to prove that, for a sufficiently large system and under sufficiently coarse observation conditions, regardless of the initial state of the system (whether it is a pure state or a mixed state), random unitary evolution will make the observational entropy approach its maximum value \(\log d\). Here \(d\) is the dimension of the Hilbert space of the system.
3. **Physically reasonable random evolution**:
- The paper further explores the use of approximate 2 - designs as a model of random evolution. Although the bounds given by this model are weaker, they are more physically reasonable because they can be efficiently implemented. Even in this case, for a sufficiently large system and under sufficiently coarse observation, the state of the system will quickly become indistinguishable from the microcanonical distribution.
### Key formulas and concepts:
- **Definition of observational entropy**:
\[
S_P(\rho)=-\sum_x\frac{\text{Tr}[P_x\rho]}{\text{Tr}[P_x]}\log\left(\frac{\text{Tr}[P_x\rho]}{\text{Tr}[P_x]}\right)=\log d - D(P(\rho)\|P(u))
\]
where \(P = \{P_x\}\) is a positive operator - valued measure (POVM), \(\rho\) is the density matrix of the system, \(u=\frac{1}{d}\) is the maximally mixed state, and \(D(\cdot\|\cdot)\) is the quantum relative entropy.
- **Expression of the Lévy - type concentration bound**:
\[
P_H\left\{S_P(U\rho U^\dagger)\leq(1 - \delta)\log d\right\}\leq\frac{4}{\kappa(P)}e^{-C\delta\kappa(P)^2d\log d}
\]
where \(\kappa(P)=\min_x\text{Tr}[P_xu]\) is an "effective roughness" parameter and \(C\approx0.0018\).
- **Concentration bound of approximate 2 - designs**:
\[
P_E\left\{S_P(U\rho U^\dagger)\leq(1 - \delta)\log d\right\}\leq\frac{4(1 + \epsilon)}{\kappa(P)^3d\log d\delta}
\]
### Conclusion:
The paper, through rigorous mathematical proofs, shows that in an isolated system, the observational entropy under