Remove the logarithmic dimension dependence for the heat maximal operator

Determine whether the logarithmic dependence on the dimension can be removed from the weak-type $(1,1)$ bound for the centered heat maximal operator $\mathcal{H}_{\ast}$ on $\mathbb{R}^n$; one possible route is to improve the Abel-power square-function estimate in Theorem 2.1 from dependence $N^{1/2}$ to $N^{1/2-\epsilon}$ for some $\epsilon>0$.

Background

The paper proves a weak-type (1,1)(1,1) estimate for the centered heat maximal operator Hf(x)=supt>0Htf(x)\mathcal{H}_{\ast}f(x)=\sup_{t>0}|H_tf(x)| with dimensional growth O(log(1+n))O(\log(1+n)). Combined with the pointwise comparison between the centered Hardy--Littlewood maximal function and the heat maximal operator, this yields an O(nlog(1+n))O(\sqrt n\log(1+n)) bound for the Hardy--Littlewood maximal function over Euclidean balls.

The logarithmic loss arises in the low Mellin-frequency estimate, which depends on a square-function bound for Abel powers whose current dimensional dependence is N\sqrt N. The authors identify improving that dependence to N1/2ϵN^{1/2-\epsilon} as one possible way to eliminate the logarithmic factor, but explicitly state that their methods do not achieve this improvement.

References

We do not know whether it is possible to remove the logarithmic dependence on dimension in our Theorem \ref{thm:main}. One approach would be to improve Theorem \ref{thm:resolvent-square} with dependence $N{1/2-\epsilon}$, though we were not able to do so with our methods.

The weak-type (1,1) bound for the Hardy--Littlewood maximal function is $O(\sqrt{n} \log n)$  (2609.05377 - Spector et al., 4 Sep 2026) in Introduction, immediately following the proof outline of Theorem 1.2