Exact value of separator exponent b_ε
Informal open question on the exact value of $b_\varepsilon$ · arXiv:2001.09679
Status open medium confidence
Dvořák establishes the bounds $\frac{1}{2\varepsilon}-1\leq b_\varepsilon\leq\frac{1}{2\varepsilon}-\frac{1}{2}$ for $0<\varepsilon<\frac{1}{2}$ and asks whether the remaining gap of $\frac{1}{2}$ can be closed; no follow-up paper resolving this question was found in five web searches. The question of whether $b_\varepsilon$ is discontinuous at $\varepsilon=1/2$ remains open.
Reviewer notes. No follow-up work resolving the gap between the upper and lower bounds on $b_\varepsilon$ was found in the indexed literature. The paper is 6 years old (2020–2026), so medium rather than high confidence is assigned. Related work on strongly sublinear separators (e.g., arXiv:2208.10074 on product structure, and a SoCG 2025 paper on sphere intersection graphs) continues the broader programme but does not appear to address this specific bound-tightening question.
Context
After establishing $\frac{1}{2\varepsilon}-1\leq b_\varepsilon\leq\frac{1}{2\varepsilon}-\frac{1}{2}$ for $0<\varepsilon<\frac{1}{2}$ and $b_\varepsilon=0$ for $\frac{1}{2}\leq\varepsilon\leq 1$, the paper notes that the remaining gap of $\frac{1}{2}$ in the bounds on $b_\varepsilon$ is unresolved, and that the behaviour at $\varepsilon=1/2$ makes the problem structurally interesting.
Notes. Stated as an open remark in the concluding summary. PDF source.
Source paper
A note on sublinear separators and expansion
Zdeněk Dvořák · 2020-07-07
https://arxiv.org/abs/2001.09679
PDF source