Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper investigates the connection between parametric Rényi divergence threshold (RDT) and ultrametric overlap gap properties (OGPs) in symmetric binary perceptrons (SBPs) to characterize statistical computational gaps. The authors derive rigorous upper bounds on the constraint densities associated with s-level ultrametric OGPs using an analytical union-bounding program, finding that the bounds at the first two levels closely approach the 3rd and 4th lifting level parametric RDT estimates. Based on these numerical results, the authors conjecture a link between ultrametric OGPs and parametric RDT, suggesting a potential isomorphism connecting their key parameters.
Forget about incremental improvements: this work hints at a deep isomorphism between two seemingly distinct geometric features of solution spaces in symmetric binary perceptrons, potentially unlocking new theoretical insights into statistical computational gaps.
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 space. Specifically, for any positive integer $s$, we consider $s$-level ultrametric OGPs ($ult_s$-OGPs) and rigorously upper-bound the associated constraint densities $α_{ult_s}$. 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, $\barα_{ult_1} \approx 1.6578$ and $\barα_{ult_2} \approx 1.6219$, closely approach the 3rd and 4th lifting level parametric RDT estimates, $α_c^{(3)} \approx 1.6576$ and $α_c^{(4)} \approx 1.6218$. 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 $ult$-OGP and parametric RDT. Specifically, we conjecture that algorithmic threshold $α_a=\lim_{s\rightarrow\infty} α_{ult_s} = \lim_{s\rightarrow\infty} \barα{ult_s} = \lim_{r\rightarrow\infty} α_{c}^{(r)}$, and $α_{ult_s} \leq α_{c}^{(s+2)}$ (with possible equality for some (maybe even all) $s$). Finally, we discuss the potential existence of a full isomorphism connecting all key parameters of $ult$-OGP and parametric RDT.