We show an area law in the mutual information for the maximally-mixed state Ω in the ground space of general Hamiltonians, which is independent of the underlying ground space degeneracy. Our result assumes the existence of a ‘good’ approximation to the ground state projector (a good AGSP), a crucial ingredient in former area-law proofs. Such approximations have been explicitly derived for 1D gapped local Hamiltonians and 2D frustration-free locally-gapped local Hamiltonians. As a corollary, we show that in 1D gapped local Hamiltonians, for any є>0 and any bi-partition L∪ Lc of the system,
We show an area law in the mutual information for the maximally-mixed state
Ω in the ground space of general Hamiltonians, which is independent of the underlying ground space degeneracy. Our result assumes the existence of a `good' approximation to the ground state projector (a good AGSP), a crucial ingredient in previous area-law proofs. Such approximations have been explicitly derived for 1D gapped local Hamiltonians and 2D frustration-free locally-gapped Hamiltonians. As a corollary, we show that in 1D gapped local Hamiltonians, for any
ε>0 and any bi-partition
L∪Lc of the system, \begin{align*} \mathrm I_{\max}^\varepsilon (L:L^c)_Ω \le \mathrm O \big( \log (|L|\log(d))+\log(1/\varepsilon)\big), \end{align*} where
∣L∣ represents the number of sites in
L,
d is the dimension of a site and $ \mathrm I_{\max}^\varepsilon (L:L^c)_Ω $ represents the
ε-\emph{smoothed maximum mutual information} with respect to the
L:Lc partition in
Ω. From this bound we then conclude
I(L:Lc)Ω≤O(log(∣L∣log(d))) -- an area law for the mutual information in 1D systems with a logarithmic correction. In addition, we show that
Ω can be approximated in trace norm up to
ε with a state of Schmidt rank of at most
poly(∣L∣/ε), leading to a good MPO approximation for
Ω with polynomial bond dimension. Similar corollaries are derived for the mutual information of 2D frustration-free and locally-gapped local Hamiltonians.
We show an area law in the mutual information for the maximally-mixed state
Ω in the ground space of general Hamiltonians, which is independent of the underlying ground space degeneracy. Our result assumes the existence of a ‘good’ approximation to the ground state projector (a good AGSP), a crucial ingredient in previous area-law proofs. Such approximations have been explicitly derived for 1D gapped local Hamiltonians and 2D frustration-free locally-gapped Hamiltonians. As a corollary, we show that in 1D gapped local Hamiltonians, for any
ϵ>0 and any bipartition
L∪Lc of the system,
Imaxϵ(L:Lc)Ω≤O(log(∣L∣log(d))+log(1/ϵ)), where |L| represents the number of sites in L, d is the dimension of a site and
Imaxϵ(L:Lc)Ω represents the
ϵ -smoothed maximum mutual information with respect to the
L:Lc partition in
Ω . From this bound we then conclude
I(L:Lc)Ω≤O(log(∣L∣log(d))) – an area law for the mutual information in 1D systems with a logarithmic correction. In addition, we show that
Ω can be approximated in trace norm up to
ϵ with a state of Schmidt rank of at most
poly(∣L∣log(d)/ϵ) , leading to a good MPO approximation for
Ω with polynomial bond dimension. Similar corollaries are derived for the mutual information of 2D frustration-free and locally-gapped local Hamiltonians.
We prove that the entanglement entropy of the ground state of a locally gapped frustration-free 2D lattice spin system satisfies an area law with respect to a vertical bipartition of the lattice into left and right regions. We first establish that the ground state projector of any locally gapped frustration-free 1D spin system can be approximated to within error є by a degree O(√nlog(є−1)) multivariate polynomial in the interaction terms of the Hamiltonian. This generalizes the optimal bound on the approximate degree of the boolean AND function, which corresponds to the special case of commuting Hamiltonian terms. For 2D spin systems we then construct an approximate ground state projector (AGSP) that employs the optimal 1D approximation in the vicinity of the boundary of the bipartition of interest. This AGSP has sufficiently low entanglement and error to establish the area law using a known technique.
An area law is proved for the Renyi entanglement entropy of possibly degenerate ground states in one-dimensional gapped quantum systems. Suppose in a chain of
n spins the ground states of a local Hamiltonian with energy gap
ϵ are constant-fold degenerate. Then, the Renyi entanglement entropy
Rα(0<α<1) of any ground state across any cut is upper bounded by
O~(α−3/ϵ), and any ground state can be well approximated by a matrix product state of subpolynomial bond dimension
2O~(ϵ−1/4log3/4n).
We give a new proof for the area law for general 1D gapped systems, which exponentially improves Hastings' famous result \cite{ref:Has07}. Specifically, we show that for a chain of d-dimensional spins, governed by a 1D local Hamiltonian with a spectral gap \eps>0, the entanglement entropy of the ground state with respect to any cut in the chain is upper bounded by $O{\frac{\log^3 d}{\eps}}$. Our approach uses the framework Arad et al to construct a Chebyshev-based AGSP (Approximate Ground Space Projection) with favorable factors. However, our construction uses the Hamiltonian directly, instead of using the Detectability lemma, which allows us to work with general (frustrated) Hamiltonians, as well as slightly improving the $1/\eps$ dependence of the bound in Arad et al. To achieve that, we establish a new, "random-walk like", bound on the entanglement rank of an arbitrary power of a 1D Hamiltonian, which might be of independent interest: \ER{H^\ell} \le (\ell d)^{O(\sqrt{\ell})}. Finally, treating d as a constant, our AGSP shows that the ground state is well approximated by a matrix product state with a sublinear bond dimension $B=e^{O(\log^{3/4}n/\eps^{1/4})}. Using this in conjunction with known dynamical programing algorithms, yields an algorithm for a 1/\poly(n) approximation of the ground energy with a subexponential running time T\le \exp(e^{O(\log^{3/4}n/\eps^{1/4})}).
We give an improved analysis of approximate ground space projectors (AGSPs) to obtain a sharp error reduction bound in the setting a degenerate ground space. The improved tools imply that the recently proven 2D area law directly extends to sub-exponentially degenerate ground spaces and also yields a sub-exponential-time classical algorithm to compute the ground states. This time complexity cannot be improved beyond sub-exponential, even for the special case of classical unfrustrated systems in 2D.