Outline of the Wang-Zahl proof of the Kakeya conjecture in ℝ 3 \mathbb{R}^{3}

Larry Guth · 2025

Read Outline of the Wang-Zahl proof of the Kakeya conjecture in ℝ 3 \mathbb{R}^{3} by Larry Guth (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-05475.txt.

Introduction

Larry Guth lays out, in outline, a recent proof of the Kakeya conjecture in three dimensions. The conjecture concerns sets that contain a unit line segment pointing in every direction, and claims such sets must fill space in a strong sense. The proof belongs to Wang and Zahl and appeared in a recent paper by them.

Guth's outline explains how the full claim is made to follow from a special sticky case of the problem. That case was settled in earlier work by Wang and Zahl, building on an approach suggested by Katz and Tao. The outline is written to make the structure of the argument visible: which parts carry the weight, and how the sticky case feeds the general one.

Catalog record

CREATORLarry Guth
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE81 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-05475

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Outline of the Wang-Zahl proof of the Kakeya conjecture in ℝ 3 \mathbb{R}^{3} Larry Guth Abstract In the recent preprint [ 5 ] , Hong Wang and Joshua Zahl proved the 3-dimensional Kakeya conjecture, building on previous papers [ 3 ] and [ 4 ] . I am working on digesting and checking the proof. I wrote an expository article [ 1 ] giving an overview of the Kakeya problem and the proof as a whole. The papers [ 3 ] and [ 4 ] prove the sticky case of the Kakeya conjecture, roughly following an approach suggested by Katz and Tao, and dealing with significant technical issues. The paper [ 5 ] reduces the general case to the sticky case. Both parts of the proof are difficult and significant, but reducing to the sticky case is the part that seemed completely out of reach until recently. In these followup notes, we give a detailed outline of how to reduce the general Kakeya conjecture to the sticky case. The argument here is a small variation on the one in [ 5 ] , which I hope simplifies some technical details. 1. Introduction to the Kakeya problem Suppose that \mathbb{T} is a set of tubes in \mathbb{R}^{3} with radius \delta and length 1. We write U(\mathbb{T})=\bigcup_{T\in\mathbb{T}}T. Our goal is to understand |U(\mathbb{T})| , the volume of U(\mathbb{T}) . We write |\mathbb{T}| for the cardinality of \mathbb{T} and we write |T| for the volume of a tube T\in\mathbb{T} . We begin with a naive question. Question 1.1 . If \mathbb{T} is a set of \delta -tubes in \mathbb{R}^{3} , is it true that |U(\mathbb{T})|\gtrapprox|\mathbb{T}||T|? The answer to this question is no. One problem is that all of the tubes of \mathbb{T} could be almost equal to each other, so that |U(\mathbb{T})| is almost |T| . To avoid this, we assume that the tubes of \mathbb{T} are essentially distinct, meaning that |T_{1}\cap T_{2}|\geq(1/2)|T_{1}| for any distinct T_{1},T_{2}\in\mathbb{T} . Throughout the survey, we will assume that we are working with essentially distinct sets of tubes. But even if we assume that the tubes of \mathbb{T} are essentially distinct, the answer to this question is still no. Here is a key counterexample. Suppose that K\subset\mathbb{R}^{3} is a convex set with dimensions a\times b\times 1 , with \delta\leq a,b . Suppose that \mathbb{T} is a maximal set of essentially distinct tubes contained in K . If |K| is much bigger than |T| , then we will have |\mathbb{T}||T|\gg|K| . In this example, the tubes \mathbb{T} cluster in the convex set K . Next we would like to consider a set of tubes that does not cluster in any convex sets. First we need some language to make this precise. Suppose that \mathbb{W} is a set of convex sets in \mathbb{R}^{n} . We will consider how the sets of \mathbb{W} cluster in various other convex sets. More generally, suppose that \mathbb{W} is a finite set of convex sets in \mathbb{R}^{n} and suppose that K is another convex set. We define (1) \mathbb{W}[K]:=\{W\in\mathbb{W}:W\subset K\} (2) \Delta(\mathbb{W},K):=\frac{\sum_{W\in\mathbb{W}[K]}|W|}{|K|} We can think of \Delta(\mathbb{W},K) as a kind of density that measures how \mathbb{W} concentrates in K . We define a maximal density (3) \Delta_{max}(\mathbb{W}):=\max_{K\textrm{ convex}}\Delta(\mathbb{W},K) The inequality \Delta_{max}(\mathbb{W})\lessapprox 1 captures the idea that \mathbb{W} does not cluster in convex sets. Theorem 1.2 . (Kakeya conjecuture, Wang-Zahl, [ 5 ] ) Suppose that \mathbb{T} is a set of \delta\times\delta\times 1 -tubes in B_{1}\subset\mathbb{R}^{3} with \Delta_{max}(\mathbb{T})\lessapprox 1 . Then |U(\mathbb{T})|\gtrapprox|\mathbb{T}||T|. Remark. The original Kakeya conjecture concerns direction-separated tubes. One version says that if \mathbb{T} is a set of \delta^{-(n-1)} \delta -tubes in \mathbb{R}^{n} with \delta -separated directions, then |U(\mathbb{T})|\gtrapprox 1 . It is straightforward to check that if the tubes of \mathbb{T} are direction separated, then \Delta_{max}(\mathbb{T})\lesssim 1 , and so Theorem 1.2 implies this form of the Kakeya conjecture in \mathbb{R}^{3} . The 2-dimensional version of the Kakeya conjecture follows from a short argument based on Cauchy-Schwarz, which has been known for many years. The 3-dimensional version of the Kakeya conjecture has been intensively investigated for many years. It is worth mentioning that the direct analogue of Theorem 1.2 is false in dimensions n\geq 4 , because of counterexamples where the tubes \mathbb{T} cluster in the \delta -neighborhood of a low degree algebraic variety. For instance, in \mathbb{R}^{4} , the tubes can cluster in the \delta -neighborhood of the quadratic hypersurface defined by x_{1}^{2}+x_{2}^{2}-x_{3}^{2}-x_{4}^{2}=1 . It is possible that Theorem 1.2 can be generalized to higher dimensions by replacing convex sets by semi-algebraic sets of small complexity. Instead of considering only the union of the tubes, it is also natural to consider a subset of each tube, which is called a shading. In general, if \mathbb{W} is a set of convex sets in \mathbb{R}^{n} , a shading Y is a subset Y(W)\subset W for each W\in\mathbb{W} . If \mathbb{W} is equipped with a shading, we write U(\mathbb{W})=\bigcup_{W\in\mathbb{W}}Y(W) . If we want to emphasize the shading, we can write U(\mathbb{W},Y) . We can generalize Theorem 1.2 to allow shadings, as long as each Y(W) is large. If \mathbb{W} has a shading Y , we define (4) \lambda(Y):=\min_{W\in\mathbb{W}}\frac{|Y(W)|}{|W|} Theorem 1.3 . (Kakeya conjecuture, Wang-Zahl, [ 5 ] ) Suppose that \mathbb{T} is a set of \delta\times\delta\times 1 -tubes in B_{1}\subset\mathbb{R}^{3} with a shading Y . If \Delta_{max}(\mathbb{T})\lessapprox 1, and \lambda(Y)\gtrapprox 1 , then |U(\mathbb{T},Y)|\gtrapprox|\mathbb{T}||T|. Remark. This small generalization of Theorem 1.2 implies the Hausdorff dimension version of the Kakeya conjecture. A Kakeya set in \mathbb{R}^{n} is a set containing a unit line segment in every direction. Theorem 1.3 implies that a Kakeya set in \mathbb{R}^{3} has Hausdorff dimension 3. The proof of Theorem 1.3 involves a complex induction, and shadings help to make the induction work, and so it is actually crucial to prove the slightly more general Theorem 1.3 . 1.1. The goal of these notes The previous papers of Wang and Zahl, [ 3 ] and [ 4 ] , prove the sticky case of the Kakeya conjecture. We will recall the precise statement below. The paper [ 5 ] gives a complex inductive argument which reduces the general case of Theorem 1.3 to the sticky case. The goal of these notes is to give a detailed outline of this reduction. We will try to explain and motivate the main ideas. We will also give a complete outline, which includes all the cases of the argument. On the other hand, we will leave out a lot of details related to pigeonholing, keeping track of shadings, and keeping track of small parameters. In Section 12 , we briefly comment on what is left out. In particular, we will use \lessapprox and \ll in an imprecise way. We write A\lessapprox B to mean “ A is not much bigger than B ” and A\ll B to mean “ A is much smaller than B ”. We outline here a small variation of the proof in [ 5 ] . We have made small adjustments to streamline the proof. Acknowledgements. Some of these adjustments were worked out in conversation with Josh and Hong. It was helpful and fun to talk about the Kakeya proof with Josh and Hong and many others at Oberwolfach in July

Read the complete text