ExplorerData ScienceMachine Learning
Research PaperResearchia:202604.22033

Ultrametric OGP - parametric RDT \emph{symmetric} binary perceptron connection

Mihailo Stojnic

Abstract

In [97,99,100], an fl-RDT framework is introduced to characterize \emph{statistical computational gaps} (SCGs). Studying \emph{symmetric binary perceptrons} (SBPs), [100] obtained an \emph{algorithmic} threshold estimate $α_a\approx α_c^{(7)}\approx 1.6093$ at the 7th lifting level (for $κ=1$ margin), closely approaching $1.58$ local entropy (LE) prediction [18]. In this paper, we further connect parametric RDT to overlap gap properties (OGPs), another key geometric feature of the solution spa...

Submitted: April 22, 2026Subjects: Machine Learning; Data Science

Description / Details

In [97,99,100], an fl-RDT framework is introduced to characterize \emph{statistical computational gaps} (SCGs). Studying \emph{symmetric binary perceptrons} (SBPs), [100] obtained an \emph{algorithmic} threshold estimate αaαc(7)1.6093α_a\approx α_c^{(7)}\approx 1.6093 at the 7th lifting level (for κ=1κ=1 margin), closely approaching 1.581.58 local entropy (LE) prediction [18]. In this paper, we further connect parametric RDT to overlap gap properties (OGPs), another key geometric feature of the solution space. Specifically, for any positive integer ss, we consider ss-level ultrametric OGPs (ultsult_s-OGPs) and rigorously upper-bound the associated constraint densities αultsα_{ult_s}. To achieve this, we develop an analytical union-bounding program consisting of combinatorial and probabilistic components. By casting the combinatorial part as a convex problem and the probabilistic part as a nested integration, we conduct numerical evaluations and obtain that the tightest bounds at the first two levels, αˉult11.6578\barα_{ult_1} \approx 1.6578 and αˉult21.6219\barα_{ult_2} \approx 1.6219, closely approach the 3rd and 4th lifting level parametric RDT estimates, αc(3)1.6576α_c^{(3)} \approx 1.6576 and αc(4)1.6218α_c^{(4)} \approx 1.6218. We also observe excellent agreement across other key parameters, including overlap values and the relative sizes of ultrametric clusters. Based on these observations, we propose several conjectures linking ultult-OGP and parametric RDT. Specifically, we conjecture that algorithmic threshold αa=limsαults=limsαˉults=limrαc(r)α_a=\lim_{s\rightarrow\infty} α_{ult_s} = \lim_{s\rightarrow\infty} \barα{ult_s} = \lim_{r\rightarrow\infty} α_{c}^{(r)}, and αultsαc(s+2)α_{ult_s} \leq α_{c}^{(s+2)} (with possible equality for some (maybe even all) ss). Finally, we discuss the potential existence of a full isomorphism connecting all key parameters of ultult-OGP and parametric RDT.


Source: arXiv:2604.19712v1 - http://arxiv.org/abs/2604.19712v1 PDF: https://arxiv.org/pdf/2604.19712v1 Original Link: http://arxiv.org/abs/2604.19712v1

Please sign in to join the discussion.

No comments yet. Be the first to share your thoughts!

Access Paper
View Source PDF
Submission Info
Date:
Apr 22, 2026
Topic:
Data Science
Area:
Machine Learning
Comments:
0
Bookmark