arXiv:2610.01987
On the Metastability of the Mean-Field Interchange Model for Local Dimension $\geq 3$
Sergio Escobar, Lin Lin, Michael Ragone, Kevin D. Stubbs
Abstract
Many quantum systems are believed to thermalize slowly at low temperatures due to the existence of metastable phases. However, there are cases where further cooling can restore polynomial-time thermalization: we demonstrate that the Davies dynamics of the mean-field interchange model with local dimension $d\geq 3$ and single-site generalized Pauli couplings exhibits this behavior. The dynamical phase diagram is beholden to two distinguished inverse temperatures, $β_{\mathrm{min}}(d)$ and $β=d$, with the static first-order phase transition lying between them. At low and high temperatures surrounding the interval $(β_{\mathrm{min}}(d),d)$, the spectral gap of the Davies generator is bounded below by an inverse polynomial in system size $n$. Meanwhile, in this interval, the gap vanishes exponentially in $n$, indicating metastability. The proof uses Schur-Weyl duality to decompose the Davies generator into a classical Markov generator on Young diagrams, and the restriction to the remaining symmetry sectors whose gap is $Ω(n^{-1})$. Thus, exponential slowdown arises entirely from the Young diagram chain.
AI Summary
The paper asks when local Davies dynamics for the mean-field interchange Hamiltonian $H=-n^{-1}\sum_{i<j}T_{ij}$ thermalizes rapidly or exhibits metastability. It considers fixed local dimension $d\geq3$ and fixed inverse temperature $\beta$, with normalized single-site Gell–Mann couplings and Metropolis rates. The spectral gap is inverse-polynomial in system size $n$ for $\beta<\beta_{\min}(d)$ and $\beta>d$, but $\exp[-\Theta_{\beta,d}(n)]$ throughout $\beta_{\min}(d)<\beta<d$. Thus further cooling restores polynomial-time mixing. Here $\beta_{\min}(d)=\min_{r>1}\frac{r+d-1}{r-1}\log r$, with $\beta_{\min}(d)=\log d+\log\log d+1+o_d(1)$. The equilibrium first-order transition occurs strictly inside this window, at $\beta_c(d)=\frac{2(d-1)}{d-2}\log(d-1)$. Schur–Weyl duality reduces the symmetry-invariant observable sector to an explicit reversible Markov chain on Young diagrams, allowing one-box moves. Its stationary probabilities are polynomially equivalent to $\exp\{n[F_\beta(\lambda/n)-\max F_\beta]\}$, where $F_\beta(x)=\frac\beta2\sum_i x_i^2-\sum_i x_i\log x_i$. Two local maxima yield exponentially small conductance inside the metastable window; controlled free-energy paths establish polynomial gap bounds outside it. Comparison with collective $SU(d)$ and permutation mixers bounds the remaining sectors by $1/n$ and a positive temperature-dependent constant, respectively, locating all exponential slowdown in the Young-diagram chain. The estimates exclude both spinodal endpoints and do not determine their critical scaling or shrinking temperature windows. Extension beyond the complete graph remains open because the exact symmetry reduction no longer closes.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 49 pages, 3 figures