Comment on “Extension of the adiabatic theorem”
Jie Gu · 2026
Read Comment on “Extension of the adiabatic theorem” by Jie Gu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-16439.txt.
Introduction
Gu offers a counterexample to a recent conjecture about the adiabatic theorem for quantum quenches within a single phase. The conjecture held that the overlap between the starting ground state and later states would peak for the later ground state. Gu constructs a local, translationally invariant, gapped free-fermion case where this fails. The before and after systems link through a symmetry-preserving gapped path, with a continuous spectrum in the large-system limit. The short note shows the claim does not hold in general.
Catalog record
| CREATOR | Jie Gu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 5421 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.16439
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Comment on “Extension of the adiabatic theorem”
Jie Gu
Abstract
In Ref. DamerowKehrein2026 , Damerow and Kehrein proposed the conjecture that, for quantum quenches within the same phase, the overlap between the initial ground state and postquench eigenstates is maximal for the postquench ground state. We show that this statement is not valid in general. An explicit local, translationally invariant, gapped free-fermion counterexample exists even though the pre- and postquench Hamiltonians are connected by a symmetry-preserving gapped path and the thermodynamic-limit spectrum is continuous.
In Ref.
DamerowKehrein2026
, the following conjecture was formulated
[Eq. (2) of that work]:
\max_{n}\left|\langle\psi_{n}^{(f)}|GS_{i}\rangle\right|^{2}=\left|\langle GS_{f}|GS_{i}\rangle\right|^{2},
(1)
for quenches between Hamiltonians in the same phase.
Here
|GS_{i}\rangle
and
|GS_{f}\rangle
denote the pre- and postquench ground states,
respectively, and
\{|\psi_{n}^{(f)}\rangle\}
are the eigenstates of the postquench
Hamiltonian.
We provide a counterexample.
Consider a one-dimensional translationally invariant two-band free-fermion model
H_{\mu}=\sum_{k}\Psi_{k}^{\dagger}h_{\mu}(k)\Psi_{k},\qquad\Psi_{k}=\begin{pmatrix}a_{k}\\
b_{k}\end{pmatrix},
(2)
with
\displaystyle h_{f}(k)
\displaystyle=\varepsilon(k)\,\sigma_{z},
(3a)
\displaystyle h_{i}(k)
\displaystyle=\varepsilon(k)\,\bigl(\cos\phi\,\sigma_{z}+\sin\phi\,\sigma_{x}\bigr),
(3b)
and
\varepsilon(k)=m+t\cos k,\qquad m>|t|>0.
(4)
Because
\varepsilon(k)
contains only a constant and a cosine harmonic, both
Hamiltonians are local nearest-neighbor models in real space.
At half filling, both
H_{i}
and
H_{f}
are gapped. Their single-particle eigenvalues are
E_{\pm}(k)=\pm\varepsilon(k),
(5)
so the gap is finite for all
k
. Moreover, the interpolation
h_{s}(k)=\varepsilon(k)\,\bigl[\cos(s\phi)\,\sigma_{z}+\sin(s\phi)\,\sigma_{x}\bigr],\qquad s\in[0,1],
(6)
preserves translation symmetry and particle-number conservation, while its spectrum
remains
\pm\varepsilon(k)
for all
s
. Hence
H_{i}
and
H_{f}
lie in the same phase
according to the definition adopted in Ref.
DamerowKehrein2026
. In the
thermodynamic limit,
k
becomes continuous, and the spectrum is correspondingly
continuous as required by the premise of Ref.
DamerowKehrein2026
.
Let
\beta_{k,\pm}^{\dagger}
create the negative- and positive-energy eigenstates of
h_{f}(k)
. The occupied negative-energy state of
h_{i}(k)
is then
d_{k}^{\dagger}=\cos\frac{\phi}{2}\,\beta_{k,-}^{\dagger}-\sin\frac{\phi}{2}\,\beta_{k,+}^{\dagger}.
(7)
For a finite system with
N
occupied momenta, the initial ground state is
|GS_{i}\rangle=\prod_{k}d_{k}^{\dagger}|0\rangle,
(8)
whereas the postquench ground state is
|GS_{f}\rangle=\prod_{k}\beta_{k,-}^{\dagger}|0\rangle.
(9)
Any postquench eigenstate at the same filling is specified by a subset
S
of momenta
promoted to the upper band:
|S\rangle=\prod_{k\in S}\beta_{k,+}^{\dagger}\prod_{k\notin S}\beta_{k,-}^{\dagger}|0\rangle.
(10)
Since different momentum sectors factorize independently, one immediately finds
\left|\langle S|GS_{i}\rangle\right|^{2}=\left(\cos^{2}\frac{\phi}{2}\right)^{N-|S|}\left(\sin^{2}\frac{\phi}{2}\right)^{|S|}.
(11)
In particular, for the postquench ground state,
\left|\langle GS_{f}|GS_{i}\rangle\right|^{2}=\left(\cos^{2}\frac{\phi}{2}\right)^{N}.
(12)
Now consider a single particle-hole excitation
|q\rangle
, corresponding to
|S|=1
.
Equation (
11
) gives
\frac{\left|\langle q|GS_{i}\rangle\right|^{2}}{\left|\langle GS_{f}|GS_{i}\rangle\right|^{2}}=\tan^{2}\frac{\phi}{2}.
(13)
Therefore, whenever
\phi>\pi/2
, one has
\left|\langle q|GS_{i}\rangle\right|^{2}>\left|\langle GS_{f}|GS_{i}\rangle\right|^{2},
(14)
so the conjecture in Eq. (
1
) is violated.
For a concrete choice, take
\phi=\frac{3\pi}{4}.
(15)
Then
\tan^{2}\frac{\phi}{2}=\tan^{2}\frac{3\pi}{8}=(1+\sqrt{2})^{2}>1,
(16)
and thus even a single excited eigenstate has larger overlap with the initial ground
state than the final ground state.
In fact, Eq. (
11
) shows more. If
\phi>\pi/2
, then
\sin^{2}\frac{\phi}{2}>\cos^{2}\frac{\phi}{2},
(17)
so the overlap increases monotonically with
|S|
. The maximal overlap is therefore
attained for the highly excited state in which all occupied momenta are promoted to the
upper band:
|S_{\max}\rangle=\prod_{k}\beta_{k,+}^{\dagger}|0\rangle,
(18)
for which
\left|\langle S_{\max}|GS_{i}\rangle\right|^{2}=\left(\sin^{2}\frac{\phi}{2}\right)^{N}>\left(\cos^{2}\frac{\phi}{2}\right)^{N}=\left|\langle GS_{f}|GS_{i}\rangle\right|^{2}.
(19)
We conclude that the conjecture proposed in Ref.
DamerowKehrein2026
does not
hold in general, even within the class of local, translationally invariant, gapped
free-fermion Hamiltonians connected by a symmetry-preserving gapped path and having a
continuous spectrum in the thermodynamic limit. The positive results obtained in
Ref.
DamerowKehrein2026
for the TFIM and special cases of the ANNNI model must
therefore rely on additional structure not captured by the conjecture in its present
form.
Note added: Subsequent to submission, we note a preprint deriving an exact necessary-and-sufficient criterion for identifying when ground-state dominance is valid or violated for translational-invariant free-fermion Hamiltonians
Haque2026