On the Last Kervaire Invariant Problem

Weinan Lin, Guozhen Wang, and Zhouli Xu · 2024

Read On the Last Kervaire Invariant Problem by Weinan Lin, Guozhen Wang, and Zhouli Xu (2024) 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-2412-10879.txt.

Introduction

Weinan Lin, Guozhen Wang, and Zhouli Xu settle, in a 2024 paper, the last open case of the Kervaire invariant problem, a question in topology that stood for decades. The problem asks in which dimensions there exist smooth framed shapes carrying a particular invariant value of one. Case by case, the answer had narrowed to a short list of dimensions and one stubborn remainder: dimension 126.

The authors prove that a certain element in the Adams spectral sequence, a central computing device of the field, survives as a permanent cycle. That single fact guarantees the shapes exist in dimension 126. Combined with earlier theorems, the full answer is now exact: such shapes exist in dimensions 2, 6, 14, 30, 62, and 126, and nowhere else. The paper closes a problem that shaped algebraic topology for generations.

Catalog record

CREATORWeinan Lin, Guozhen Wang, and Zhouli Xu
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE174 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-10879

Browse the full library · About the Tower of Babel

Full text

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

On the Last Kervaire Invariant Problem Weinan Lin, Guozhen Wang, and Zhouli Xu Abstract . We prove that the element h 6 2 h_{6}^{2} is a permanent cycle in the Adams spectral sequence. As a result, we establish the existence of smooth framed manifolds with Kervaire invariant one in dimension 126, thereby resolving the final case of the Kervaire invariant problem. Combining this result with the theorems of Browder, Mahowald–Tangora, Barratt–Jones–Mahowald, and Hill–Hopkins–Ravenel, we conclude that smooth framed manifolds with Kervaire invariant one exist in and only in dimensions 2 , 6 , 14 , 30 , 62 2,6,14,30,62 , and 126 126 . 1. Introduction For a smooth framed manifold in dimension 4k+2 , the Kervaire invariant, taking values in \mathbb{F}_{2} , determines whether the manifold could be converted into a homotopy sphere via surgery – it takes value 0 if the manifold can be converted to a homotopy sphere and 1 otherwise. Formally speaking, the Kervaire invariant is defined as the Arf invariant of the quadratic refinement of the intersection pairing in the cohomology of the manifold with \mathbb{F}_{2} coefficients. Using this invariant, Kervaire [ 33 ] discovered a PL-manifold in dimension 10 that does not admit any smooth structure. The Kervaire invariant problem seeks to identify the dimensions in which there exist framed smooth manifolds with Kervaire invariant one. In these dimensions, exactly half of the cobordism classes of framed manifolds have Kervaire invariant one, and the other half have Kervaire invariant 0. The Kervaire invariant problem is closely related to many problems in differential topology, especially Kervaire–Milnor’s classification theorem [ 34 ] on exotic smooth structures on spheres. In this paper we prove the following Theorem 1.1 . Theorem 1.1 . There exist framed manifolds of Kervaire invariant one in dimension 126. Together with previous results by Browder [ 10 ] , Mahowald–Tangora [ 44 ] , Barratt–Jones–Mahowald [ 4 ] , and Hill–Hopkins–Ravenel [ 23 ] , this is the last case of the Kervaire invariant problem. Corollary 1.2 . The dimensions that there exist framed manifolds of Kervaire invariant one are 2, 6, 14, 30, 62, and 126. In dimensions 2, 6, and 14, the product of spheres S^{1}\times S^{1},\ S^{3}\times S^{3},S^{7}\times S^{7} can be framed to have Kervarie invariant one. In dimension 30, an explicit framed manifold of Kervarie invariant one was constructed by J.Jones in [ 32 ] . For dimensions 62 and 126, we would like to comment that no explicit manifold of Kervarie invariant one was known, although 50\% of all framed manifolds in these two dimensions have Kervarie invariant one. Our Theorem 1.1 is a consequence of Browder’s theorem (Theorem 1.3 ) and the following main theorem of this paper, Theorem 1.4 . Theorem 1.3 (Browder [ 10 ] ) . There exist framed manifolds in dimension n of Kervaire invariant one if and only if (1) n=2^{j+1}-2 , and (2) the element h_{j}^{2} in the Adams spectral sequence survives to the E_{\infty} -page. The Adams spectral sequence [ 1 ] converges to the stable homotopy groups of spheres, which, by Pontryagin’s theorem, correspond to framed manifolds up to framed cobordism. When the element h_{j}^{2} survives (see below for an introduction to the Adams spectral sequence and the element h_{j}^{2} ), the homotopy classes it detects are denoted by \theta_{j}\in\pi_{2^{j+1}-2} . Browder showed that these classes correspond to framed manifolds with Kervaire invariant one. Theorem 1.4 (Theorem 7.1 ) . The element h_{6}^{2} survives to the E_{\infty} -page in the Adams spectral sequence. In addition to Theorem 1.1 , further implications of Theorem 1.4 , particularly in manifold topology and unstable homotopy theory, can be found in [ 47 , 42 , 23 , 3 ] , among others. Recall that the 2-primary Adams spectral sequence has the following form. E_{2}^{s,t}=\mathrm{Ext}^{s,t}_{A}(\mathbb{F}_{2},\mathbb{F}_{2})\Longrightarrow\pi_{t-s}{S^{0}}, d_{r}:E_{r}^{s,t}\longrightarrow E_{r}^{s+r,t+r-1}. Here the E_{2} -page is the cohomology of the mod 2 Steenrod algebra A, S^{0} is the 2-completed sphere spectrum, and the spectral sequence converges to the 2-completed stable homotopy groups of spheres. In general, for a spectrum X , its \mathrm{H}{\mathbb{F}}_{2} -Adams spectral sequence is denoted by E_{r}^{*,*}(X) , converging to the 2-completed homotopy groups of X . For the sphere, Adams [ 1 ] computed the Adams 1 -line: \mathrm{Ext}^{1,t}_{A}(\mathbb{F}_{2},\mathbb{F}_{2})=\begin{cases}\mathbb{F}_{2},\ \text{if}\ t=2^{j}\ \text{for some}\ j\geq 0,\\ 0,\ \text{otherwise}.\end{cases} Denote by h_{j} the generator of \mathbb{F}_{2} in the bidegree (s,t)=(1,2^{j}) . Adams [ 2 ] proved that the element h_{j} survives in the Adams spectral sequence if and only if j\geq 3 , resolving the famous Hopf invariant problem. In fact, Adams proved that d_{2}(h_{j})=h_{0}h_{j-1}^{2}\neq 0,\ \text{for}\ j\geq 4. The Kervaire invariant element h_{j}^{2} lives on the Adams 2-line, which is generated by elements of the form \{h_{i}h_{j}\} that are subject to the relations h_{i}h_{i+1}=0 . Notably, h_{6}^{2} was the last unknown element on the Adams 2-line whose survival in the Adams spectral sequence remained uncertain prior to our work. As a corollary of Theorem 1.4 , we have: Corollary 1.5 . On the Adams 2-line, the only non-trivial elements that survive in the Adams spectral sequence are: h_{0}h_{2},\ h_{0}h_{3},\ h_{2}h_{4},\ h_{1}h_{j}\ (j\geq 3),\ h_{j}^{2}\ (j\leq 6). Proof. From Adams’s Hopf invariant one differentials, Lin’s computations of the Adams 4-line [ 40 ] , and the Leibniz rule, we know that the only elements that survive to the Adams E_{3} -page are h_{0}h_{2},\ h_{0}h_{3},\ h_{2}h_{4},\ h_{2}h_{5},\ h_{3}h_{5},\ h_{3}h_{6},\ h_{1}h_{j}\ (j\geq 3),\ h_{j}^{2}\ (j\leq 6). May [ 45 ] proved that the first three elements and h_{j}^{2} for j\leq 3 survive. Mahowald–Tangora [ 44 ] proved that h_{2}h_{5} supports a nonzero d_{3} -differential, h_{3}h_{5} supports a nonzero d_{4} -differential, and h_{4}^{2} survives. Isaksen–Wang–Xu [ 29 ] proved that h_{3}h_{6} supports a nonzero d_{3} -differential. Mahowald [ 43 ] proved that h_{1}h_{j} for j\geq 3 survive. Barratt–Jones–Mahowald [ 4 ] proved that h_{5}^{2} survives (see [ 56 , 29 ] for alternative proofs). Hill–Hopkins–Ravenel [ 23 ] proved that h_{j}^{2} supports nonzero differentials for j\geq 7 (the targets of these differentials are still unknown). Finally, by Theorem 1.4 , h_{6}^{2} survives. ∎ Before Hill–Hopkins–Ravenel’s proof on the non-existence of \theta_{j} for j\geq 7 , Barratt–Jones–Mahowald [ 3 ] had an inductive argument trying to establish the existence of all \theta_{j} : They proved that if there exists a \theta_{j} satisfying 2\cdot\theta_{j}=0,\ \theta_{j}^{2}=0 , then there exists a \theta_{j+1} satisfying 2\cdot\theta_{j+1}=0 . In particular, the problem of whether a \theta_{j+1} of order 2 exists is known as the strong Kervaire invariant problem. It was known that 2\cdot\theta_{j}=0 for j\leq 5 (see [ 44 ] for j=4 and [ 56 , 29 ] for j=5 ). Although we prove that \theta_{6} exists in this paper, the following questions are still open: Question 1.6 . Does there exist a \theta_{6} that has order 2? Question 1.7 . Does there exist a \theta_{5} such that \theta_{5}^{2}=0 ? By Barratt–Jones–Mahowald’s theorem [ 3 ] , if the answer to Question 1.7 is positive, then it would imply our Theorem 1.4 and a positive answer to Question 1.6 . We plan to study these questions in a future project. As suggested by the proof of Corollary 1.5 , computation of differentials in the Adams spectral sequence has a long history. See Section 2

Read the complete text