ABC implies that Ramanujan’s tau function misses almost all primes

David Kurniadi Angdinata †, Evan Chen †, Chris Cummins *, Ben Eltschig *, Dejan Grubisic *, Leopold Haller *, Letong Hong ⋄ \diamond, Andranik Kurghinyan *, Kenny Lau *, Hugh Leather *, Seewoo Lee †, Simon Mahns *, Aram H. Markosyan *, Rithikesh Muddana *, Ken Ono †, Manooshree Patel *, Gaurang Pendharkar *, Vedant Rathi *, Alex Schneidman *, Volker Seeker *, Shubho Sengupta ⋄ \diamond, Ishan Sinha *, Jimmy Xin *, Jujian Zhang †* · 2026

Read ABC implies that Ramanujan’s tau function misses almost all primes by David Kurniadi Angdinata †, Evan Chen †, Chris Cummins *, Ben Eltschig *, Dejan Grubisic *, Leopold Haller *, Letong Hong ⋄ \diamond, Andranik Kurghinyan *, Kenny Lau *, Hugh Leather *, Seewoo Lee †, Simon Mahns *, Aram H. Markosyan *, Rithikesh Muddana *, Ken Ono †, Manooshree Patel *, Gaurang Pendharkar *, Vedant Rathi *, Alex Schneidman *, Volker Seeker *, Shubho Sengupta ⋄ \diamond, Ishan Sinha *, Jimmy Xin *, Jujian Zhang †* (2026) 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-2603-29970.txt.

Introduction

Angdinata and co-authors bound how often Ramanujan's tau function hits prime values. A folklore guess says infinitely many primes appear as its absolute values. Assuming the abc conjecture, they prove a much stronger upper bound. The count of such primes up to X grows at most like X to the 13/22 power. So tau misses almost all primes, a density one set. A heuristic still predicts infinitely many hits. The proof's engine was formalized in Lean by AxiomProver.

Catalog record

CREATORDavid Kurniadi Angdinata †, Evan Chen †, Chris Cummins *, Ben Eltschig *, Dejan Grubisic *, Leopold Haller *, Letong Hong ⋄ \diamond, Andranik Kurghinyan *, Kenny Lau *, Hugh Leather *, Seewoo Lee †, Simon Mahns *, Aram H. Markosyan *, Rithikesh Muddana *, Ken Ono †, Manooshree Patel *, Gaurang Pendharkar *, Vedant Rathi *, Alex Schneidman *, Volker Seeker *, Shubho Sengupta ⋄ \diamond, Ishan Sinha *, Jimmy Xin *, Jujian Zhang †*
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE23963
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.29970

Browse the full library · About the Tower of Babel

Full text

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

ABC implies that Ramanujan’s tau function misses almost all primes David Kurniadi Angdinata †, Evan Chen †, Chris Cummins *, Ben Eltschig *, Dejan Grubisic *, Leopold Haller *, Letong Hong ⋄ \diamond, Andranik Kurghinyan *, Kenny Lau *, Hugh Leather *, Seewoo Lee †, Simon Mahns *, Aram H. Markosyan *, Rithikesh Muddana *, Ken Ono †, Manooshree Patel *, Gaurang Pendharkar *, Vedant Rathi *, Alex Schneidman *, Volker Seeker *, Shubho Sengupta ⋄ \diamond, Ishan Sinha *, Jimmy Xin *, Jujian Zhang †* Abstract . Lehmer conjectured that Ramanujan’s tau-function never vanishes. In a related direction, a folklore conjecture asserts that infinitely many primes arise as absolute values of Ramanujan’s tau-function. Recently, Xiong showed that these prime values form a subset of the primes with density at most 2 / 11 2/11 . Assuming the a ​ b ​ c abc Conjecture, we prove the stronger upper bound S ⁡ ( X ) := # ⁡ { ℓ ≤ X : ℓ ​ prime and ​ | τ ⁡ ( n ) | = ℓ ​ for some ​ n ≥ 1 } = O ⁡ ( X 9 / 10 ​ log ⁡ X ) , S(X):=\#\{\ell\leq X:\ \ell\ \text{prime and }|\tau(n)|=\ell\text{ for some }n\geq 1\}=O(X^{9/10}\log X), which implies that Ramanujan’s tau-function misses a density 1 subset of the primes. We give a heuristic suggesting that S ⁡ ( X ) S(X) should nevertheless be infinite, with predicted order of magnitude S ⁡ ( X ) ≍ C ​ X 1 11 ( log ⁡ X ) 2 . S(X)\asymp\frac{CX^{\frac{1}{11}}}{(\log X)^{2}}. The main engine in this note was formalized and produced automatically in Lean/Mathlib by AxiomProver from a natural-language statement of the problem. 2020 Mathematics Subject Classification Primary 11F11; Secondary 11F30 Authors are listed alphabetically. † Mathematical contributor, * Engineering contributor, \diamond Principal investigator. 1. Introduction and statement of results Ramanujan’s discriminant modular form (1.1) \Delta(z)=q\prod_{n\geq 1}(1-q^{n})^{24}=\sum_{n\geq 1}\tau(n)q^{n},\qquad(q=e^{2\pi iz}), is the normalized weight 12 cusp form on \mathrm{SL}_{2}(\mathbb{Z}) with integer Fourier coefficients \tau(n) [ 21 ] . Lehmer famously conjectured that \tau(n)\neq 0 for all n\geq 1 [ 12 ] , and he also initiated the study of primality in the image of \tau [ 13 ] . Serre proved that the primes p with \tau(p)=0 have natural density 0 [ 22 ] . This result was further refined by Thorner and Zaman [ 23 ] . Beyond nonvanishing, one may ask which integers lie in the image of \tau . For fixed odd integers \alpha , Murty–Murty–Shorey proved that \tau(n)=\alpha has only finitely many solutions [ 19 ] . In recent years, there has been substantial interest in finding explicit omitted values of \tau . A foundational structural input for many such results is a prime-power criterion of Balakrishnan–Craig–Tsai and one of the present authors [ 3 ] . Using this criterion, they [ 3 ] proved, for n\geq 2 , that \tau(n)\notin\{\pm 1,\pm 3,\pm 5,\pm 7,\pm 13,\pm 17,-19,\pm 23,\pm 37,\pm 691\}. After that, many authors have used this criterion to find additional omitted values (for example, see [ 1 , 9 , 10 ] ). Furthermore, Bennett–Gherga–Patel–Siksek [ 6 ] proved that \tau(n)\neq\pm\ell^{m} for all odd primes 3\leq\ell<100 and all integers m\geq 1 . Regarding even values, Balakrishnan–Ono–Tsai [ 4 ] produced the first explicit even integers known not to occur as \tau -values. A folklore conjecture (for example, see [ 6 , 15 ] ) asserts that the \tau -values include infinitely many primes up to sign. In the complementary direction, Xiong [ 24 ] bounded (from above) the density of such values. Namely, he investigated the function (1.2) S(X):=\#\{\ell\leq X:\ \ell\text{ prime with }|\tau(n)|=\ell\ \text{ for some }n\}. He proved that in 18 of the non-zero residue classes 1 1 1 The obstruction pertains to the primes in the four congruence classes \ell\equiv 1,3,5,22\pmod{23}. modulo 23 , prime \tau -values form a thin set with \lim_{X\rightarrow+\infty}\frac{S(X)}{\pi(X)}\leq\frac{2}{11}, where \pi(X) denotes the number of primes \leq X. In this note, we address the primes in these classes conditionally. We let (1.3) \mathcal{P}:=\{\pm\ell:\ \ell\text{ is an odd prime}\}\qquad{\text{and}}\qquad\mathcal{V}:=\{\tau(n):n\geq 1\}. To study \mathcal{P}\cap\mathcal{V} , the odd primes (up to sign) that occur as \tau -values, we employ the celebrated abc Conjecture of Masser and Oesterlé [ 20 , Exp. 694] . Conjecture ( abc Conjecture) . For every \varepsilon>0, there exists a constant C_{\varepsilon}>0 such that for all coprime nonzero integers a,b,c with a+b=c one has |c|\leq C_{\varepsilon}\,\operatorname{rad}(abc)^{1+\varepsilon}, where \operatorname{rad}(abc):=\prod_{p\mid abc}p is the product of distinct primes dividing abc. Assuming this conjecture, we obtain the following improvement to Xiong’s theorem. Theorem 1 . Assuming the abc Conjecture, as X\rightarrow+\infty we have S(X)=O(X^{\frac{9}{10}}\log X). Remark 2 . Theorem 1 implies that Ramanujan’s tau-function misses almost all primes under the abc Conjecture. In other words, we have \frac{S(X)}{\pi(X)}\longrightarrow 0, which follows by the Prime Number Theorem \lim_{X\rightarrow+\infty}\frac{\pi(X)}{X/\log X}=1. For completeness, we point out that there are prime values of Ramanujan’s tau-function. Indeed, Lehmer [ 13 ] found that \tau(251^{2})=-80561663527802406257321747. Remark 3 . Theorem 1 holds in much greater generality. The same conclusion holds for the coefficients of any normalized Hecke eigenform on \mathrm{SL}_{2}(\mathbb{Z}) with integer coefficients. To see that, one notes that the same Hecke relations hold, which means that the proof of Theorem 1.2 of [ 24 ] applies mutatis mutandis . One then uses the fact that every integer Hecke eigenvalue \lambda(p) , for odd primes p in level 1, is even (for example, see [ 11 ] ). This implies the direct analog of Proposition 8 in this note. Assuming the abc Conjecture, Theorem 1 establishes that the \tau -values can only include a density zero subset of the primes (up to sign). In Section 4 , we address the conjecture that there are still infinitely many such values. We employ the truth of the Sato-Tate Conjecture to suggest that S(X)\asymp\frac{CX^{\frac{1}{11}}}{(\log X)^{2}}. This paper is organized as follows. In Section 2 , we prove the main engine of the paper, which are two analytic estimates for the number of integer points that lie near two specific hyperelliptic curves under the abc Conjecture. In Section 3 , we then use these estimates to prove Theorem 1 . As mentioned above, in Section 4 we suggest an analytic estimate for S(X) . Finally, in Section 5 , we describe the protocol we employed to formalize and automatically generate the two estimates from a natural-language statement of the problem. 2. The main engine To prove Theorem 1 , we make use of the following proposition which is a straightforward consequence of Xiong’s Theorem (see the proof of Theorem 1.2 of [ 24 ] ). Lemma 4 . For the functions \begin{split}E_{2}(X)&:=\#\{(x,y)\in\mathbb{Z}_{\geq 1}\times\mathbb{Z}:\ x>X^{2/11},\ 1\leq|x^{11}-y^{2}|\leq X\},\\ E_{4}(X)&:=\#\{(x,u)\in\mathbb{Z}_{\geq 1}\times\mathbb{Z}:\ x>X^{1/11},\ 1\leq|5x^{22}-u^{2}|\leq 4X\},\end{split} as X\rightarrow+\infty we have S(X)\ \ll\ X^{9/10}\log X\ +\ X^{13/22}\ +\ X^{6/11}\ +\ E_{2}(X)\ +\ E_{4}(X). We defer the proof of this lemma to the next section. Theorem 1 follows from the next result, the main engine of this note. Lemma 5 . Assume the abc Conjecture. (i) For every \eta>0 , we have that E_{2}(X)\ll_{\eta}X^{4/9+\eta}. (ii) For every \eta>0 , we have that E_{4}(X)\ll_{\eta}X^{1/5+\eta}. 2.1. The prime-power reduction We begin with the prime-power reduction which underlies the

Read the complete text