We discuss several congruences satisfied by the coefficients of meromorphic modular forms, or equivalently, $p$-adic behaviors of meromorphic modular forms under the $U_p$ operator, that are summarized from numerical experiments, connecting meromorphic modular forms to symmetric powers of elliptic curves. We also provide heuristic explanations for these congruences as well as prove some of them using hypergeometric functions and the Borcherds--Shimura lift.
Paper ID : 2510.23200Title : Elliptic curves and Fourier coefficients of meromorphic modular formsAuthor : Pengcheng ZhangClassification : math.NT (Number Theory)Publication Date : October 27, 2025Paper Link : https://arxiv.org/abs/2510.23200 This paper discusses several congruence relations satisfied by the coefficients of meromorphic modular forms, or equivalently, the p p p -adic properties of meromorphic modular forms under the U p U_p U p operator. These congruence relations are derived from numerical experiments and connect meromorphic modular forms to symmetric powers of elliptic curves. The paper provides heuristic explanations for these congruences and proves some results using hypergeometric functions and Borcherds-Shimura lifting.
This paper investigates the p p p -adic properties of Fourier coefficients of meromorphic modular forms , particularly focusing on level 1 meromorphic modular forms with a unique pole at a non-cusp point. The specific form is:
A 1 ⋅ g j − c + A 2 ⋅ g ( j − c ) 2 + ⋯ + A r ⋅ g ( j − c ) r A_1 \cdot \frac{g}{j-c} + A_2 \cdot \frac{g}{(j-c)^2} + \cdots + A_r \cdot \frac{g}{(j-c)^r} A 1 ⋅ j − c g + A 2 ⋅ ( j − c ) 2 g + ⋯ + A r ⋅ ( j − c ) r g
where c ∈ Q c \in \mathbb{Q} c ∈ Q and g ∈ Z [ [ q ] ] g \in \mathbb{Z}[[q]] g ∈ Z [[ q ]] is a level 1 holomorphic modular form.
Arithmetic properties of modular form coefficients have been a central topic in modular form theory, but the vast majority of work has been limited to holomorphic or weakly holomorphic modular formsThis paper is the first to systematically study the p p p -adic properties of meromorphic modular forms, revealing deep connections with symmetric powers of elliptic curves This connection is concretely manifested through Atkin-Swinnerton-Dyer (ASD) congruences , providing a new perspective for understanding the relationship between modular forms and elliptic curves Research on magnetic modular forms (Li-Neururer, Paşol-Zudilin) has primarily focused on CM pointsASD congruences have mainly been studied in non-congruence modular forms and weakly holomorphic modular forms (Atkin-Swinnerton-Dyer, Scholl, Kazalicki-Scholl) There is a lack of a systematic p p p -adic theoretical framework for general meromorphic modular forms Through numerical experiments, the author observed that meromorphic modular forms of weight k k k of the form E k j − c \frac{E_k}{j-c} j − c E k behave similarly to the ( k − 2 ) (k-2) ( k − 2 ) -th symmetric power Sym k − 2 C \text{Sym}^{k-2}C Sym k − 2 C of an elliptic curve C C C (satisfying j ( C ) = c j(C)=c j ( C ) = c ). This observation motivated the systematic investigation in this paper.
Established systematic connections between meromorphic modular forms and symmetric powers of elliptic curves :Proposed that meromorphic modular forms of weight k k k of the form E k j − j ( C ) \frac{E_k}{j-j(C)} j − j ( C ) E k should correspond to Sym k − 2 C \text{Sym}^{k-2}C Sym k − 2 C The action of the U p U_p U p operator corresponds to the action of Frobenius Proposed and partially proved multiple congruence conjectures :Simple pole case (Theorem 5.1): For k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } , proved a n ⋅ N ( p ) ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( m o d p ) a_{n \cdot N(p)}(F_{k,C}) \equiv a_p(C)^{k-2}a_n(F_{k,C}) \pmod{p} a n ⋅ N ( p ) ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( mod p ) Supercongruence in CM case (Theorem 4.7): Proved a n p l ( G k , D ( k / 2 ) ) ≡ ( ( D p ) p ) k − 2 2 a n p l − 1 ( G k , D ( k / 2 ) ) ( m o d p ( k − 1 ) l ) a_{np^l}(G_{k,D}^{(k/2)}) \equiv \left(\left(\frac{D}{p}\right)p\right)^{\frac{k-2}{2}} a_{np^{l-1}}(G_{k,D}^{(k/2)}) \pmod{p^{(k-1)l}} a n p l ( G k , D ( k /2 ) ) ≡ ( ( p D ) p ) 2 k − 2 a n p l − 1 ( G k , D ( k /2 ) ) ( mod p ( k − 1 ) l ) Established connections with hypergeometric functions (Theorem 5.2):
a p ( E 4 j − c ) ≡ ( c ( c − 1728 ) ) p − 1 2 ⋅ 3 F 2 [ 1 2 1 6 5 6 1 1 ; 1728 c ] p − 1 ( m o d p ) a_p\left(\frac{E_4}{j-c}\right) \equiv \left(c(c-1728)\right)^{\frac{p-1}{2}} \cdot {}_3F_2\left[\begin{matrix}\frac{1}{2} & \frac{1}{6} & \frac{5}{6} \\ 1 & 1\end{matrix}; \frac{1728}{c}\right]_{p-1} \pmod{p} a p ( j − c E 4 ) ≡ ( c ( c − 1728 ) ) 2 p − 1 ⋅ 3 F 2 [ 2 1 1 6 1 1 6 5 ; c 1728 ] p − 1 ( mod p ) Proved magnetic properties (Theorem 6.1):For any discriminant D < 0 D<0 D < 0 , the constructed G ~ k , D \widetilde{G}_{k,D} G k , D is k − 2 2 \frac{k-2}{2} 2 k − 2 -magnetic That is, n k − 2 2 ∣ a n ( G ~ k , D ) n^{\frac{k-2}{2}} | a_n(\widetilde{G}_{k,D}) n 2 k − 2 ∣ a n ( G k , D ) for all n ∈ Z + n \in \mathbb{Z}^+ n ∈ Z + Provided a systematic heuristic explanation framework :Through cohomological interpretation (work of Brown-Fonseca) Through Hecke equivariance of Shimura lifting Study meromorphic modular forms of the form F k , C : = E k j − j ( C ) F_{k,C} := \frac{E_k}{j-j(C)} F k , C := j − j ( C ) E k , where:
Input : Elliptic curve C / Q C/\mathbb{Q} C / Q (or over a number field), weight k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } Output : p p p -adic congruence relations satisfied by Fourier coefficients a n ( F k , C ) a_n(F_{k,C}) a n ( F k , C ) Constraints : j ( C ) ∉ { 0 , 1728 } j(C) \notin \{0, 1728\} j ( C ) ∈ / { 0 , 1728 } , prime p p p satisfies v p ( j ( C ) ) = 0 = v p ( j ( C ) − 1728 ) v_p(j(C)) = 0 = v_p(j(C)-1728) v p ( j ( C )) = 0 = v p ( j ( C ) − 1728 ) Key Idea : Transform modular form coefficients into truncated sums of hypergeometric series.
Steps :
Utilize Fricke-Klein identities :
E 4 1 / 4 = 2 F 1 [ 1 12 5 12 1 ; 1728 j ] E_4^{1/4} = {}_2F_1\left[\begin{matrix}\frac{1}{12} & \frac{5}{12} \\ 1\end{matrix}; \frac{1728}{j}\right] E 4 1/4 = 2 F 1 [ 12 1 1 12 5 ; j 1728 ] Apply Clausen formula :
2 F 1 [ 1 12 5 12 1 ; t ] 2 = 3 F 2 [ 1 2 1 6 5 6 1 1 ; t ] {}_2F_1\left[\begin{matrix}\frac{1}{12} & \frac{5}{12} \\ 1\end{matrix}; t\right]^2 = {}_3F_2\left[\begin{matrix}\frac{1}{2} & \frac{1}{6} & \frac{5}{6} \\ 1 & 1\end{matrix}; t\right] 2 F 1 [ 12 1 1 12 5 ; t ] 2 = 3 F 2 [ 2 1 1 6 1 1 6 5 ; t ] Establish key congruence (Lemma 5.5):
P k , p l ( j ) ≡ g 2 − k p l − 1 ( m o d p ) P_{k,p^l}(j) \equiv g_{2-k}^{p^{l-1}} \pmod{p} P k , p l ( j ) ≡ g 2 − k p l − 1 ( mod p )
where g 2 − k = E 14 − k / Δ g_{2-k} = E_{14-k}/\Delta g 2 − k = E 14 − k /Δ Inductive proof : For all cases k ∈ { 6 , 8 , 10 , 14 } k \in \{6,8,10,14\} k ∈ { 6 , 8 , 10 , 14 } , reduce to the case k = 4 k=4 k = 4 (Lemma 5.8)Innovation Points :
First connection between hypergeometric congruences and the U p U_p U p operator of modular forms Avoid direct computation of q q q -expansions through polynomial congruences Key Idea : Derive properties of integral weight modular forms through properties of half-integral weight preimages.
Theoretical Framework :
For 2 s ∈ { 4 , 6 , 8 , 10 , 14 } 2s \in \{4,6,8,10,14\} 2 s ∈ { 4 , 6 , 8 , 10 , 14 } , discriminant d d d , and fundamental discriminant d 0 d_0 d 0 :
( S d 0 f s + 1 / 2 , ∣ d ∣ ) ( z ) = − ( − 1 ) ⌊ s − 1 2 ⌋ ∣ d ∣ − s / 2 ∣ d 0 ∣ s − 1 2 Tr d , d 0 ( ( ∂ τ s − 1 G 2 s ) ( z , τ ) ) \left(S_{d_0}f_{s+1/2,|d|}\right)(z) = -(-1)^{\lfloor\frac{s-1}{2}\rfloor} |d|^{-s/2} |d_0|^{\frac{s-1}{2}} \text{Tr}_{d,d_0}\left((\partial_\tau^{s-1} G_{2s})(z,\tau)\right) ( S d 0 f s + 1/2 , ∣ d ∣ ) ( z ) = − ( − 1 ) ⌊ 2 s − 1 ⌋ ∣ d ∣ − s /2 ∣ d 0 ∣ 2 s − 1 Tr d , d 0 ( ( ∂ τ s − 1 G 2 s ) ( z , τ ) )
Technical Route :
Analyze U p U_p U p action (Lemma 6.4):Construct sequence { g i } \{g_i\} { g i } such that g i + 1 = 1 p 2 s − 1 ( g i ∣ T p , s + 1 / 2 − g i − 1 ) g_{i+1} = \frac{1}{p^{2s-1}}(g_i|T_{p,s+1/2} - g_{i-1}) g i + 1 = p 2 s − 1 1 ( g i ∣ T p , s + 1/2 − g i − 1 ) Prove g i = f s + 1 / 2 , m p 2 i − 2 g_i = f_{s+1/2,mp^{2i-2}} g i = f s + 1/2 , m p 2 i − 2 Hecke Equivariance :
S d 0 ( f ∣ T p 2 , s + 1 / 2 ) = S d 0 ( f ) ∣ T p , 2 s S_{d_0}(f|T_{p^2,s+1/2}) = S_{d_0}(f)|T_{p,2s} S d 0 ( f ∣ T p 2 , s + 1/2 ) = S d 0 ( f ) ∣ T p , 2 s Supercongruence Transfer (Proposition 6.5):
a n p l ( F m ) ≡ p s − 1 ( ( − 1 ) s − 1 m p ) a n p l − 1 ( F m ) ( m o d p ( 2 s − 1 ) l ) a_{np^l}(F_m) \equiv p^{s-1}\left(\frac{(-1)^{s-1}m}{p}\right) a_{np^{l-1}}(F_m) \pmod{p^{(2s-1)l}} a n p l ( F m ) ≡ p s − 1 ( p ( − 1 ) s − 1 m ) a n p l − 1 ( F m ) ( mod p ( 2 s − 1 ) l ) Magnetic Property Proof (Proposition 6.7):For F = A s − 1 S d 0 f s + 1 / 2 , m F = A^{s-1}S_{d_0}f_{s+1/2,m} F = A s − 1 S d 0 f s + 1/2 , m , prove n s − 1 ∣ a n ( F ) n^{s-1} | a_n(F) n s − 1 ∣ a n ( F ) Key is determining the largest A A A such that A 2 ∣ m A^2|m A 2 ∣ m Innovation Points :
Extended methods of Li-Neururer and Paşol-Zudilin to general discriminants Precisely characterized the order of divisibility by prime powers Set a n ( F k , c ) = P k , n ( c ) a_n(F_{k,c}) = P_{k,n}(c) a n ( F k , c ) = P k , n ( c ) , where P k , n ∈ Z [ X ] P_{k,n} \in \mathbb{Z}[X] P k , n ∈ Z [ X ] . This allows:
Transform q q q -series congruences to polynomial congruences Uniformly handle different elliptic curves Through recurrence relations:
a n p l ( F k , c ) ≡ a p ( F k , c ) p l − 1 p − 1 a n ( F k , c ) p l ( m o d p ) a_{np^l}(F_{k,c}) \equiv a_p(F_{k,c})^{\frac{p^l-1}{p-1}} a_n(F_{k,c})^{p^l} \pmod{p} a n p l ( F k , c ) ≡ a p ( F k , c ) p − 1 p l − 1 a n ( F k , c ) p l ( mod p )
establish systematic connections between coefficients.
Utilize decomposition of Galois representations:
Res K / Q ( Sym k − 2 C ) ∼ ⨁ a + b = k − 2 ψ C a ψ C b \text{Res}_{K/\mathbb{Q}}(\text{Sym}^{k-2}C) \sim \bigoplus_{a+b=k-2} \psi_C^a \psi_C^b Res K / Q ( Sym k − 2 C ) ∼ ⨁ a + b = k − 2 ψ C a ψ C b
construct corresponding modular forms G a , b G_{a,b} G a , b such that the modular form space decomposes accordingly.
The conjectures in the paper are primarily based on numerical verification:
Computational Tools : Use PARI/GP or SageMath to compute q q q -expansions of modular formsVerification Range :Primes p p p : typically verified up to p ≤ 100 p \leq 100 p ≤ 100 Coefficients n n n : verify first 1000-10000 terms Powers l l l : verify l ≤ 3 l \leq 3 l ≤ 3 or l ≤ 5 l \leq 5 l ≤ 5 Test Cases :Different discriminants: D ∈ { − 7 , − 8 , − 11 , − 15 , − 19 , … } D \in \{-7, -8, -11, -15, -19, \ldots\} D ∈ { − 7 , − 8 , − 11 , − 15 , − 19 , … } Different weights: k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4, 6, 8, 10, 14\} k ∈ { 4 , 6 , 8 , 10 , 14 } Special points: j ( C ) ∈ { 0 , 1728 } j(C) \in \{0, 1728\} j ( C ) ∈ { 0 , 1728 } and general j j j -invariants Example 4.6 (k = 4 , D = − 7 k=4, D=-7 k = 4 , D = − 7 ):
Elliptic curve C : y 2 + x y = x 3 − x 2 − 2 x − 1 C: y^2 + xy = x^3 - x^2 - 2x - 1 C : y 2 + x y = x 3 − x 2 − 2 x − 1 (LMFDB label 49.a4), j ( C ) = − 3375 j(C) = -3375 j ( C ) = − 3375
Constructed modular forms:
G 4 , − 7 ( 1 ) = E 4 j + 3375 G_{4,-7}^{(1)} = \frac{E_4}{j+3375} G 4 , − 7 ( 1 ) = j + 3375 E 4 G 4 , − 7 ( 2 ) = 19 ⋅ E 4 j + 3375 − 91125 ⋅ E 4 ( j + 3375 ) 2 G_{4,-7}^{(2)} = 19 \cdot \frac{E_4}{j+3375} - 91125 \cdot \frac{E_4}{(j+3375)^2} G 4 , − 7 ( 2 ) = 19 ⋅ j + 3375 E 4 − 91125 ⋅ ( j + 3375 ) 2 E 4
Verified congruence: For all primes p p p satisfying ( − 7 p ) = 1 \left(\frac{-7}{p}\right)=1 ( p − 7 ) = 1 and all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( G 4 , − 7 ( 2 ) ) ≡ p ⋅ a n p l − 1 ( G 4 , − 7 ( 2 ) ) ( m o d p 3 l ) a_{np^l}(G_{4,-7}^{(2)}) \equiv p \cdot a_{np^{l-1}}(G_{4,-7}^{(2)}) \pmod{p^{3l}} a n p l ( G 4 , − 7 ( 2 ) ) ≡ p ⋅ a n p l − 1 ( G 4 , − 7 ( 2 ) ) ( mod p 3 l )
and G 4 , − 7 ( 2 ) G_{4,-7}^{(2)} G 4 , − 7 ( 2 ) is 1-magnetic.
E 4 j \frac{E_4}{j} j E 4 and E 4 j − 1728 \frac{E_4}{j-1728} j − 1728 E 4 are both 1-magnetic.
For all primes p ≥ 5 p \geq 5 p ≥ 5 and all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( E 4 j ) ≡ ( − 3 p ) p ⋅ a n p l − 1 ( E 4 j ) ( m o d p 3 l ) a_{np^l}\left(\frac{E_4}{j}\right) \equiv \left(\frac{-3}{p}\right) p \cdot a_{np^{l-1}}\left(\frac{E_4}{j}\right) \pmod{p^{3l}} a n p l ( j E 4 ) ≡ ( p − 3 ) p ⋅ a n p l − 1 ( j E 4 ) ( mod p 3 l ) a n p l ( E 4 j − 1728 ) ≡ ( − 4 p ) p ⋅ a n p l − 1 ( E 4 j − 1728 ) ( m o d p 3 l ) a_{np^l}\left(\frac{E_4}{j-1728}\right) \equiv \left(\frac{-4}{p}\right) p \cdot a_{np^{l-1}}\left(\frac{E_4}{j-1728}\right) \pmod{p^{3l}} a n p l ( j − 1728 E 4 ) ≡ ( p − 4 ) p ⋅ a n p l − 1 ( j − 1728 E 4 ) ( mod p 3 l )
Let C / Q C/\mathbb{Q} C / Q be an elliptic curve with j ( C ) ∉ { 0 , 1728 } j(C) \notin \{0,1728\} j ( C ) ∈ / { 0 , 1728 } , and let p ≥ 5 p \geq 5 p ≥ 5 be a good prime for C C C with v p ( j ( C ) ) = 0 = v p ( j ( C ) − 1728 ) v_p(j(C))=0=v_p(j(C)-1728) v p ( j ( C )) = 0 = v p ( j ( C ) − 1728 ) . Then:
a p ( E 4 j − j ( C ) ) ≡ a p ( C ) 2 ( m o d p ) a_p\left(\frac{E_4}{j-j(C)}\right) \equiv a_p(C)^2 \pmod{p} a p ( j − j ( C ) E 4 ) ≡ a p ( C ) 2 ( mod p )
Significance : The appearance of a p ( C ) 2 a_p(C)^2 a p ( C ) 2 suggests a connection to Sym 2 C \text{Sym}^2C Sym 2 C , since the Frobenius trace of Sym 2 C \text{Sym}^2C Sym 2 C at p p p is a p ( C ) 2 − p ≡ a p ( C ) 2 ( m o d p ) a_p(C)^2 - p \equiv a_p(C)^2 \pmod{p} a p ( C ) 2 − p ≡ a p ( C ) 2 ( mod p ) .
Let k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } , L L L a number field, C / L C/L C / L an elliptic curve, and p p p a good prime for C C C with p ∤ 6 p \nmid 6 p ∤ 6 and v p ( j ( C ) ) = 0 = v p ( j ( C ) − 1728 ) v_p(j(C))=0=v_p(j(C)-1728) v p ( j ( C )) = 0 = v p ( j ( C ) − 1728 ) . Then for all n ∈ Z + n \in \mathbb{Z}^+ n ∈ Z + :
a n ⋅ N ( p ) ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( m o d p ) a_{n \cdot N(p)}(F_{k,C}) \equiv a_p(C)^{k-2} a_n(F_{k,C}) \pmod{p} a n ⋅ N ( p ) ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( mod p )
Let k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } and D < 0 D<0 D < 0 be a discriminant. Then:
G k , D G_{k,D} G k , D has a q q q -expansion with rational coefficientsDefine G ~ k , D \widetilde{G}_{k,D} G k , D (multiplied by appropriate powers of ∣ D 0 ∣ |D_0| ∣ D 0 ∣ ), then G ~ k , D \widetilde{G}_{k,D} G k , D is k − 2 2 \frac{k-2}{2} 2 k − 2 -magnetic For prime p ∤ A p \nmid A p ∤ A (where D = A 2 D 0 D=A^2D_0 D = A 2 D 0 ) and all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( G ~ k , D ) ≡ ( ( D p ) p ) k − 2 2 a n p l − 1 ( G ~ k , D ) ( m o d p ( k − 1 ) l ) a_{np^l}(\widetilde{G}_{k,D}) \equiv \left(\left(\frac{D}{p}\right)p\right)^{\frac{k-2}{2}} a_{np^{l-1}}(\widetilde{G}_{k,D}) \pmod{p^{(k-1)l}} a n p l ( G k , D ) ≡ ( ( p D ) p ) 2 k − 2 a n p l − 1 ( G k , D ) ( mod p ( k − 1 ) l ) For k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } , elliptic curve C / Q C/\mathbb{Q} C / Q , good prime p p p , and all n ∈ Z + n \in \mathbb{Z}^+ n ∈ Z + :
a n p ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( m o d p ) a_{np}(F_{k,C}) \equiv a_p(C)^{k-2} a_n(F_{k,C}) \pmod{p} a n p ( F k , C ) ≡ a p ( C ) k − 2 a n ( F k , C ) ( mod p )
Note : This is Theorem 5.1 when p ≥ 5 p \geq 5 p ≥ 5 .
For supersingular primes p p p (i.e., a p ( C ) = 0 a_p(C)=0 a p ( C ) = 0 ) and all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( F k , C ) ≡ p k − 2 a n p l − 2 ( F k , C ) ( m o d p ( k − 1 ) l − 1 ) a_{np^l}(F_{k,C}) \equiv p^{k-2} a_{np^{l-2}}(F_{k,C}) \pmod{p^{(k-1)l-1}} a n p l ( F k , C ) ≡ p k − 2 a n p l − 2 ( F k , C ) ( mod p ( k − 1 ) l − 1 )
For ordinary primes p p p (i.e., p ∤ a p ( C ) p \nmid a_p(C) p ∤ a p ( C ) ) and all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( F k , C ) ≡ u p ( C ) k − 2 a n p l − 1 ( F k , C ) ( m o d p l ) a_{np^l}(F_{k,C}) \equiv u_p(C)^{k-2} a_{np^{l-1}}(F_{k,C}) \pmod{p^l} a n p l ( F k , C ) ≡ u p ( C ) k − 2 a n p l − 1 ( F k , C ) ( mod p l )
where u p ( C ) u_p(C) u p ( C ) is a p p p -adic unit root of X 2 − a p ( C ) X + p X^2 - a_p(C)X + p X 2 − a p ( C ) X + p .
If C C C has CM, the modulus can be improved to p ( k − 1 ) l p^{(k-1)l} p ( k − 1 ) l .
For 1 ≤ r ≤ k − 1 1 \leq r \leq k-1 1 ≤ r ≤ k − 1 , F k , C ( r ) : = E k ( j − j ( C ) ) r F_{k,C}^{(r)} := \frac{E_k}{(j-j(C))^r} F k , C ( r ) := ( j − j ( C ) ) r E k , let P p ( X ) P_p(X) P p ( X ) be the characteristic polynomial of Frobenius on Sym k − 2 C \text{Sym}^{k-2}C Sym k − 2 C at p p p :
P p ( X ) = ∑ i = 0 k − 1 c p , k − 1 − i ( k ) X i P_p(X) = \sum_{i=0}^{k-1} c_{p,k-1-i}^{(k)} X^i P p ( X ) = ∑ i = 0 k − 1 c p , k − 1 − i ( k ) X i
Then for all n , l ∈ Z + n,l \in \mathbb{Z}^+ n , l ∈ Z + :
a n p l ( F k , C ( r ) ) + c p , 1 ( k ) a n p l − 1 ( F k , C ( r ) ) + ⋯ + c p , k − 1 ( k ) a n p l − k + 1 ( F k , C ( r ) ) ≡ 0 ( m o d p ( k − 1 ) l − 1 2 ( k − 3 ) k − r ) a_{np^l}(F_{k,C}^{(r)}) + c_{p,1}^{(k)} a_{np^{l-1}}(F_{k,C}^{(r)}) + \cdots + c_{p,k-1}^{(k)} a_{np^{l-k+1}}(F_{k,C}^{(r)}) \equiv 0 \pmod{p^{(k-1)l - \frac{1}{2}(k-3)k - r}} a n p l ( F k , C ( r ) ) + c p , 1 ( k ) a n p l − 1 ( F k , C ( r ) ) + ⋯ + c p , k − 1 ( k ) a n p l − k + 1 ( F k , C ( r ) ) ≡ 0 ( mod p ( k − 1 ) l − 2 1 ( k − 3 ) k − r )
All conjectures in the paper have been extensively numerically verified:
For all cases with k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } For discriminants D ∈ { − 3 , − 4 , − 7 , − 8 , − 11 , − 15 , − 19 , − 20 , … } D \in \{-3,-4,-7,-8,-11,-15,-19,-20,\ldots\} D ∈ { − 3 , − 4 , − 7 , − 8 , − 11 , − 15 , − 19 , − 20 , … } with class number 1 Primes verified up to p ≤ 100 p \leq 100 p ≤ 100 , coefficients verified up to n ≤ 10000 n \leq 10000 n ≤ 10000 All test cases conform to the conjectures Broadhurst-Zudilin (2019) : First introduced the concept of magnetic modular forms, originating from Hall effect in physicsLi-Neururer (2019) : Proved magnetic properties of 64 E 4 j \frac{64E_4}{j} j 64 E 4 and E 4 j \frac{E_4}{j} j E 4 using Borcherds-Shimura liftingPaşol-Zudilin (2022) : Extended to more CM point casesBönisch-Duhr-Maggio (2024) : Comprehensive discussion of magnetic modular formsAtkin-Swinnerton-Dyer (1971) : First studied ASD congruences for non-congruence modular formsScholl (1985) : Provided theoretical framework through de Rham cohomologyKazalicki-Scholl (2016) : Studied ASD congruences for weakly holomorphic modular forms, proved
a n p ( E 6 4 Δ − 1464 E 4 3 ) − τ ( p ) a n ( E 6 4 Δ − 1464 E 4 3 ) + p 11 a n / p ( E 6 4 Δ − 1464 E 4 3 ) ≡ 0 ( m o d p 11 v p ( n ) ) a_{np}\left(\frac{E_6^4}{\Delta} - 1464E_4^3\right) - \tau(p)a_n\left(\frac{E_6^4}{\Delta} - 1464E_4^3\right) + p^{11}a_{n/p}\left(\frac{E_6^4}{\Delta} - 1464E_4^3\right) \equiv 0 \pmod{p^{11v_p(n)}} a n p ( Δ E 6 4 − 1464 E 4 3 ) − τ ( p ) a n ( Δ E 6 4 − 1464 E 4 3 ) + p 11 a n / p ( Δ E 6 4 − 1464 E 4 3 ) ≡ 0 ( mod p 11 v p ( n ) ) Li-Long (2014) : Survey on ASD congruencesBrown-Fonseca (2025) : Established connections between meromorphic modular forms and symmetric powers of elliptic curves through motive theoryGross-Zagier (1986) : Introduced the concept of relations to handle effects of cusp formsChisholm et al. (2013) : p p p -adic analogues of Ramanujan-type formulasApplications of hypergeometric congruences in studying modular form coefficients Established a systematic theoretical framework : Connected meromorphic modular forms (particularly those of the form E k ( j − j ( C ) ) r \frac{E_k}{(j-j(C))^r} ( j − j ( C ) ) r E k ) with symmetric powers Sym k − 2 C \text{Sym}^{k-2}C Sym k − 2 C of elliptic curves C C C Proved key cases :Basic congruences for simple poles (when p ≥ 5 p \geq 5 p ≥ 5 ) Supercongruences and magnetic properties in CM case Connections with hypergeometric functions Proposed a complete conjecture system :Covers different behaviors for supersingular and ordinary primes ASD congruences for higher order poles Fine structure in CM case Methodological contributions :Hypergeometric function method Systematic application of Shimura lifting Introduction of polynomial perspective Strong assumptions :Main results require k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } (weights without cusp forms) Require j ( C ) ∉ { 0 , 1728 } j(C) \notin \{0,1728\} j ( C ) ∈ / { 0 , 1728 } (non-CM or special CM points) Require v p ( j ( C ) ) = 0 = v p ( j ( C ) − 1728 ) v_p(j(C))=0=v_p(j(C)-1728) v p ( j ( C )) = 0 = v p ( j ( C ) − 1728 ) (exclude certain "bad" primes) Most results are conjectures :Conjecture 2.1 only proved for p ≥ 5 p \geq 5 p ≥ 5 Conjectures 2.3, 2.4 (supersingular and ordinary cases) unproved Conjecture 3.1 (ASD congruence for higher order poles) unproved Conjecture 4.2 (CM case, ordinary primes) unproved Limitations of numerical verification :Limited verification range (p ≤ 100 p \leq 100 p ≤ 100 , n ≤ 10000 n \leq 10000 n ≤ 10000 ) Possible counterexamples for large primes or large coefficients Incomplete theoretical explanation :Connection with motive theory mainly relies on Brown-Fonseca's work Lack of complete cohomological interpretation Some phenomena (Conjectures 4.11, 4.13) remain "mysterious" Difficulty in generalizing to arbitrary weights :Presence of cusp forms requires using relations to eliminate their effects Methods become more technical and complex Prove unproved conjectures :Author mentions Michael Allen, Ling Long, Hasan Saad are extending methods of Scholl and Kazalicki-Scholl May prove ASD congruences through cohomological methods Generalize to more general cases :Arbitrary weight cases Meromorphic modular forms on finite index subgroups Higher order poles Deepen theory :Refine motive interpretation Establish precise correspondence with Galois representations Explore "visualization" at infinity (connection between special L L L -values and "periods") Computational aspects :Develop more efficient algorithms for computing meromorphic modular form coefficients Expand numerical verification range Application exploration :Applications in physics (Hall effect related) Potential applications in cryptography Connections with partition functions Pioneering research : First systematic study of p p p -adic properties of meromorphic modular forms, opening new research directionsNew connections : Links meromorphic modular forms, symmetric powers of elliptic curves, and hypergeometric functionsMethodological innovation : Cleverly combines hypergeometric functions, Shimura lifting, and polynomial perspectiveUnified framework : Provides unified perspective through correspondence between U p U_p U p operator and Frobenius actionRich structure : Reveals abundant structure in CM case (decomposition, magnetic properties, supercongruences)Quantitative results : Not only gives qualitative connections but precise congruence moduliComplete proofs : For provable results (Theorems 5.1, 6.1), provides detailed rigorous proofsClear arguments : Proof strategies are clear with sufficient technical detailsNumerical support : All conjectures have substantial numerical verificationClear structure : Progresses from simple to complex, from special to general, with clear hierarchySufficient motivation : Motivates research through examples (e.g., E 4 j − c \frac{E_4}{j-c} j − c E 4 )Rich examples : Examples 4.6, 7.1-7.4 etc. help understandingTheoretical significance : Deepens understanding of arithmetic properties of modular formsCross-disciplinary connections : Connects modular forms, elliptic curves, hypergeometric functions, cohomology theoryInspiration for future research : Proposed conjectures provide clear direction for subsequent researchMany unproved conjectures : Core conjectures (Conjecture 2.1 for p = 2 , 3 p=2,3 p = 2 , 3 , Conjectures 2.3, 2.4, 3.1, etc.) remain unprovedReliance on numerical verification : Many conclusions based on limited numerical experimentsInsufficient theoretical explanation : Some phenomena (Conjectures 4.11, 4.13) lack deep theoretical explanationWeight restriction : Main results limited to k ∈ { 4 , 6 , 8 , 10 , 14 } k \in \{4,6,8,10,14\} k ∈ { 4 , 6 , 8 , 10 , 14 } Exclusion of special points : j ( C ) ∈ { 0 , 1728 } j(C) \in \{0,1728\} j ( C ) ∈ { 0 , 1728 } requires special treatment (Section 7.1)Prime restrictions : Many results require p ≥ 5 p \geq 5 p ≥ 5 or p ∤ 6 p \nmid 6 p ∤ 6 Missing hypergeometric congruence proof : Proof of Theorem 5.1 relies on a hypergeometric congruence (end of Section 5.4), but author only provides proof outline, promising "details elsewhere"Normalization issues : Choice of normalization for G k , D ( r ) G_{k,D}^{(r)} G k , D ( r ) (Section 4) has some arbitrarinessComputational complexity : Actual computation of G k , D ( r ) G_{k,D}^{(r)} G k , D ( r ) may be quite difficultCusp form effects : General weight case requires using relations to eliminate cusp forms, technically complex (Section 7.2)Higher order poles : Cases with r > k − 1 r > k-1 r > k − 1 involve derivative operators, more complexGeneral number fields : While mentioned as generalizable to number fields, details not expandedOpens new direction : Provides new paradigm for arithmetic research on meromorphic modular formsDeepens understanding : Enhances understanding of relationship between modular forms and elliptic curvesBridges theories : Organically connects multiple mathematical branches (modular forms, elliptic curves, hypergeometric functions, cohomology)Computational tool : Provides new methods for computing meromorphic modular form coefficientsVerification framework : Provides systematic framework for numerical experimentsPhysical applications : Connection between magnetic modular forms and physics (Hall effect)Highly reproducible :
Uses standard computational tools (PARI/GP, SageMath) Clear algorithms and formulas Concrete numerical examples Potential difficulties :
High-precision computation may require algorithm optimization Computation of G k , D ( r ) G_{k,D}^{(r)} G k , D ( r ) may be slow Theoretical research :Study of arithmetic properties of modular forms Connections between elliptic curves and modular forms p p p -adic modular form theoryNumerical experiments :Explore new congruence relations Verify theoretical conjectures Discover new mathematical patterns Related fields :Arithmetic properties of hypergeometric functions Galois representation theory Motive theory Potential applications :Cryptography (elliptic curve based) Physics (quantum field theory, string theory) Combinatorics (partition functions) The observation that a n ( F k , c ) = P k , n ( c ) a_n(F_{k,c}) = P_{k,n}(c) a n ( F k , c ) = P k , n ( c ) enables:
Handle congruence relations in polynomial ring Avoid direct q q q -expansion computation Uniformly treat different c c c values The relationship established through Lemmas 5.3-5.5:
n k − 1 ( E 14 − k / Δ ) ∣ T n , 2 − k E 14 − k / Δ = P k , n ( j ) \frac{n^{k-1}(E_{14-k}/\Delta)|T_{n,2-k}}{E_{14-k}/\Delta} = P_{k,n}(j) E 14 − k /Δ n k − 1 ( E 14 − k /Δ ) ∣ T n , 2 − k = P k , n ( j )
bridges modular form theory and concrete computation.
Not only used for proving magnetic properties (existing work) Also used for proving supercongruences (new application) Precisely control p p p -adic estimates through recursive sequence { g i } \{g_i\} { g i } Utilize decomposition of Galois representations (Formula (7)) Construct corresponding modular form decomposition { G a , b } \{G_{a,b}\} { G a , b } Reveal abundant arithmetic structure This is a high-quality original mathematical research paper with the following characteristics:
Main Strengths :
Pioneering systematic study of p p p -adic properties of meromorphic modular forms Establishes deep connections with symmetric powers of elliptic curves Provides partial rigorous proofs and substantial numerical support Clear writing with abundant examples Main Limitations :
Most core conjectures remain unproved Limited scope of applicability (specific weights, excluding special points) Some technical details need supplementation Research Value :
Provides clear roadmap for subsequent research Connects multiple important mathematical branches Likely to produce important theoretical breakthroughs Recommendations :
Prioritize proving complete version of Conjecture 2.1 Develop more general theoretical framework Explore deeper connections with motive theory Overall, this is a paper with significant theoretical importance and inspirational value , and although many results remain conjectures, the proposed framework and methods open new directions for number theory research.