Maximally-stable Local Optima in Random Graphs and Spin Glasses: Phase Transitions and Universality
Yatin Dandi, David Gamarnik, Lenka Zdeborov\'a

TL;DR
This paper investigates phase transitions in the stability of local optima in random graphs and spin glasses, establishing universal thresholds and extending prior results to general stability levels.
Contribution
It proves the existence of a universal phase transition for $h$-stability in random graphs and spin glasses, extending previous results from $h=0$ to general $h$ values.
Findings
Existence of a phase transition at a universal $h^*$ value.
Below $h^*$, most spins are $h$-stable or $h$-friendly.
Above $h^*$, the number of unstable spins becomes linear.
Abstract
We consider -stable local optima of Ising spin glass models, defined as spin configurations such that for nearly all of the spins, flipping their values results in increasing energy by at least a given amount . Spins satisfying this condition are referred to as -stable spins for that configuration. Similarly, we consider a very related notion of -friendly partitions of a graph. These are defined as bi-partitionings such that for most nodes, the normalized number of neighbors within the node's partition exceed the normalized number of neighbors outside the partition by a certain amount . For spin glasses as well as sparse and dense random graphs, while restricting to bisections, we prove the existence of a phase transition for the normalized energy level around a universal value . For below the phase transition value , bisections exist where the number of…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTheoretical and Computational Physics · Markov Chains and Monte Carlo Methods · Complex Network Analysis Techniques
