We investigate how feature-space dimensionality reshapes relaxation and correlation structure in a transverse-field Ising model with feature-conditioned, ferromagnetically biased bond disorder, analyzed via its anisotropic (two plus one)-dimensional classical counterpart. Using Suzuki–Trotter simulations of the mapped classical system, we study imaginary-time autocorrelators, coarse equal-time spatial summaries, and eigenvalue spectra of equal-time correlation matrices. Each spin is assigned a $d$-component feature vector, and a similarity scale $\hat{\sigma }$ controls how feature differences bias the distribution of bond signs. Increasing $d$ homogenizes pairwise feature distances and, together with $\hat{\sigma }$, tunes bond-sign heterogeneity within this positively biased ensemble. We find a systematic dynamical crossover from short-$\tau$ transients at low $d$ to multistep relaxation with long-lived plateaus at higher $d$, accompanied by a compression of equal-time correlation weight into fewer dominant spatial modes. Feature-conditioned disorder thus provides a controlled framework for studying how correlated disorder reshapes relaxation spectra in disordered Ising networks, with potential relevance to programmable Ising hardware.

Close

PrevNext

View figure in article

• Adiabatic quantum optimization
• Optimization problems
• Disordered systems
• Spin glasses
• Ising model
• Quantum Monte Carlo