Exact cost of the Euclidean Maximum Traveling Salesman Problem in one dimension

Sergio Caracciolo; Carlo Vanoni

Home/Archive/arXiv

arXiv · 2610.26536

Exact cost of the Euclidean Maximum Traveling Salesman Problem in one dimension

Sergio Caracciolo·Carlo Vanoni

View original source ↗Read source document ↗Export metadata ↓

Abstract

We study the random Euclidean Maximum Traveling Salesman Problem: finding a Hamiltonian cycle of maximal total cost through any set of $N$ cities on a line, where each edge cost is a function of its Euclidean length. We characterize the maximizing cycles for convex increasing, concave increasing, and convex decreasing cost functions. For points sampled independently and uniformly from the unit interval, we derive explicit expressions for the average maximal cost for arbitrary $N$ and asymptotic expansions with power-law edge costs.

Explore related subjects

cond-mat.dis-nn cond-mat.stat-mech math-ph math.CO math.MP

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Sergio Caracciolo, Carlo Vanoni. 2026-10-19. Exact cost of the Euclidean Maximum Traveling Salesman Problem in one dimension. https://arxiv.org/abs/2610.26536

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Numerical and theoretical studies of the Anderson transition in the chiral symmetry classes with weak topology

Combining lattice-model simulations with a renormalization group (RG) analysis of effective theories, we investigate the nature of the Anderson transition in chiral symmetry classes with one-dimensional (1D) weak topology. In the simulation study, we extend previous transfer matrix analyses to the chiral symplectic class, and study numerical Lyapunov exponents via a finite-size scaling (FSS) analysis that assumes spatially isotropic scaling. The analysis shows that, as in the other two chiral symmetry classes, the weak topology induces an intermediate quasi-localized (QL) phase between metal and Anderson insulator phases. In this QL phase, the localization length of wave functions diverges exclusively along the direction of the 1D weak topology. A comparison of the universal critical exponents across the three chiral symmetry classes indicates that the quantum criticality of the metal-to-QL transition varies, depending on the nature of the time-reversal symmetry. Motivated by this numerical observation, we revisit and extend our previous two-dimensional (2D) RG analysis to all three chiral classes, now newly incorporating a one-loop renormalization of the weak topological term in the analysis. The revised analysis reveals that a quasi-localized strong-coupling fixed point previously reported in the chiral unitary class is unstable under this new inclusion; instead, the strong-coupling phase is entirely governed by a stable fixed point with conventional localized character. Nevertheless, in the chiral unitary and chiral symplectic classes, the RG analysis still yields the hallmark of the 1D weak topology through the spatially anisotropic scaling of the Anderson transition criticality. These theoretical findings suggest that a conclusive numerical identification of the 2D QL phase requires a finite-size scaling approach that accommodates generic (anisotropic) spatial scaling.

cond-mat.dis-nn↗

arXiv2026

Toward a Physics of Deep Learning and Brains

Deep neural networks and brains both learn and share superficial similarities: processing nodes are likened to neurons and adjustable weights are likened to modifiable synapses. But can a unified theoretical framework be found to underlie them both? Here we show that the equations used to describe neuronal avalanches in living brains can also be applied to cascades of activity in deep neural networks. These equations are derived from non-equilibrium statistical physics and show that deep neural networks learn best when poised between absorbing and active phases. Because these networks are strongly driven by inputs, however, they do not operate at a true critical point but within a quasi-critical regime -- one that still approximately satisfies crackling noise scaling relations. By training networks with different initializations, we show that maximal susceptibility is a more reliable predictor of learning than proximity to the critical point itself. This provides a blueprint for engineering improved network performance. Finally, using finite-size scaling we identify distinct universality classes, including Barkhausen noise and directed percolation. This theoretical framework demonstrates that universal features are shared by both biological and artificial neural networks.

cond-mat.dis-nn↗

arXiv2026

Synchronous and asynchronous Monte Carlo dynamics of mean-field spin glasses with rotationally invariant couplings: what they share and where they differ

We compare two Markov chain Monte Carlo dynamics for Ising spin glasses with rotationally invariant couplings: the synchronous Monte Carlo method (SyncMC), which updates all spins in parallel with the help of Gaussian auxiliary fields, and asynchronous Metropolis-Hastings dynamics, which updates one spin at a time. Using the Guionnet-Maïda formula, we derive the dynamical mean-field theory of both dynamics in terms of the R-transform of the coupling spectrum. The two effective single-spin processes share a retarded self-interaction and a Gaussian noise; SyncMC has, in addition, an instantaneous self-coupling and a noise that depends only on the response function. These additional terms act only on the fastest time scale, so that the equations for slow time scales, including the fluctuation-dissipation ratio and, for the Sherrington-Kirkpatrick model, the Parisi equation, are the same for the two dynamics. Simulations of the random orthogonal and Sherrington-Kirkpatrick models confirm that both dynamics approach the same fluctuation-dissipation relations, differing only in how many updates they need to do so.

cond-mat.dis-nn↗