REVIEW 1 major objections 5 minor 2 cited by
An elliptic surface with infinitely many fibers for which the rank does not jump
T0 review · 1 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash
Pith's one-line read The elliptic curve $y^2 = x(x^2 - x + T)$ over $\mathbb{Q}(T)$ is proved to have rank 0, and infinitely many rational specializations also have rank 0, giving the first unconditional instance of Conjecture 1.1.
desk verdict First unconditional example of rank-non-jumping fibers for a nonisotrivial elliptic surface; the 2-descent is solid and the only soft spot is a normalization detail in Lemma 2.5 that is correct but should be spelled out. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The argument is carried by a 2-descent through the cyclic 2-isogeny $\phi: E \to E'$, whose kernel is generated by the 2-torsion point $(0,0)$. The $\phi$-Selmer group is computed directly from the homogeneous spaces $C_d$ and their local solubility, and has order 2. The dual Selmer group $\mathrm{Sel}_{\hat{\phi}}(E'/\mathbb{Q})$ is controlled by the Schaefer-Stoll formula, which expresses $|\mathrm{Sel}_{\hat{\phi}}|/|\mathrm{Sel}_{\phi}|$ as the product of the period ratio $\Omega_E/\Omega_{E'}$, a Tamagawa-number ratio, and a torsion ratio; the period ratio is forced to be $1/2$ by a theorem on local invariants of isogenous elliptic curves, because the alternative 1 would contradict the computed root number $W(E) = 1$. An exact sequence then bounds $\mathrm{Sel}_2(E/\mathbb{Q})$ by $\mathrm{Sel}_{\hat{\phi}}(E'/\mathbb{Q})$, giving $|\mathrm{Sel}_2| \leq 2$. The infinite supply of suitable parameters $t$ comes from a theorem on 3-term arithmetic progressions in the primes congruent to 3 mod 8.
What would settle it
Compute the 2-Selmer group of the curve $E_t$ for $t = 11/6$, obtained from the prime triple $(3, 11, 19)$; the proof predicts $|\mathrm{Sel}_2| = 2$ and rank 0. If a rational point of infinite order exists, or if the Selmer group has order divisible by 4, then Lemma 2.5's period-ratio application is wrong and Theorem 1.2 fails.
Extended reading notes
Core claim
The central claim, Theorem 1.2, is that the elliptic curve $E$ over $\mathbb{Q}(T)$ given by $y^2 = x(x^2 - x + T)$ has $E(\mathbb{Q}(T))$ of rank 0, and that $E_t(\mathbb{Q})$ has rank 0 for infinitely many $t \in \mathbb{Q}$. For any prime triple $(m, m+n, m+2n)$ all congruent to 3 mod 8, the paper sets $t = (m+n)/(2m)$ and shows that the specialized curve is isomorphic to $y^2 = x(x^2 - 4m^2x + 8m^3(m+n))$. The rank bound is obtained by computing the 2-Selmer group through the cyclic 2-isogeny $\phi: E \to E'$, using the Schaefer-Stoll formula and a period-ratio theorem for isogenous elliptic curves to obtain $|\mathrm{Sel}_2(E/\mathbb{Q})| \leq 2$; since $E(\mathbb{Q})$ has a point of order 2, the rank must be 0. The infinitude of the prime triples, given by Green's theorem on 3-term arithmetic progressions in primes, makes the set of such $t$ infinite.
Load-bearing premise
The rank-zero conclusion rests on the period-ratio theorem for isogenous elliptic curves being exactly applicable, so that $\Omega_E/\Omega_{E'} = 1/2$ and never 1 for any constructed curve; if the ratio ever equaled 1, the Selmer bound $|\mathrm{Sel}_{\hat{\phi}}(E'/\mathbb{Q})| = 2$ would fail and the rank could be positive.
Editorial extensions
If this is right
- Conjecture 1.1 holds for the specific surface $E$, giving the first unconditional confirmation of that conjecture.
- For every prime triple $m, m+n, m+2n \equiv 3 \pmod{8}$, the fiber $E_t$ at $t = (m+n)/(2m)$ has 2-Selmer group of order exactly 2 and rank exactly 0.
- Silverman's specialization bound, together with infinitely many rank-0 fibers, determines the generic rank of $E(\mathbb{Q}(T))$ to be 0.
- The method of confining bad reduction to a few explicit primes, then running a 2-descent, is carried over to a rank-2 example in a follow-up paper, so the approach is not restricted to rank 0.
Reading between the lines
- One could look for other residue classes $a \bmod 8$ and other signature choices where the relevant Legendre symbols give the same period-ratio and root-number configuration; this would produce further unconditional families satisfying Conjecture 1.1.
- The proof uses the period ratio only through its possible values $\{1, 1/2\}$; a natural numerical experiment is to compute $\Omega_E/\Omega_{E'}$ for the first few prime triples to see the ratio is always $1/2$, which would test the single most delicate step of the descent.
- A quantitative version of the argument might supply an explicit positive-density statement for the rank-0 fibers in this family, since the prime triples from Roth-type theorems are numerous; the paper itself only needs infinitude.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper constructs a nonisotrivial elliptic surface E/Q(T) defined by y^2 = x(x^2 - x + T) and proves Theorem 1.2: the group E(Q(T)) has rank 0, and there are infinitely many t in Q for which the specialized curve E_t has rank 0. The proof specializes at t = (m+n)/(2m), where m, m+n, m+2n are primes congruent to 3 mod 8; Green's theorem on 3-term arithmetic progressions in primes supplies infinitely many such triples. For these t, the curve E_t is isomorphic to the curve E: y^2 = x(x^2 - 4m^2 x + 8m^3(m+n)). The paper then performs a 2-descent on E, introducing a 2-isogenous curve E' and computing the associated Selmer groups. The local solvability computation (Lemma 2.4) shows |Sel_phi(E/Q)| = 2, and a Cassels period-ratio argument (Lemma 2.5) shows |Sel_phihat(E'/Q)| = 2. The exact sequence relating these Selmer groups to Sel_2(E/Q) then gives |Sel_2(E/Q)| <= 2, which forces the rank of E(Q) to be 0.
Significance. If correct, this is the first unconditional example of Conjecture 1.1 for a nonisotrivial elliptic curve over Q(T): infinitely many fibers have rank equal to the generic rank (here 0), rather than the generic rank being exceeded. The proof is largely self-contained: the 2-descent is spelled out in detail, the local root numbers and Tamagawa numbers are obtained via Tate's algorithm and cited tables, and the only external inputs are Green's theorem, Cassels' formula, and the Dokchitser-Dokchitser results. The construction is not tuned to the conclusion; the same descent works for every admissible triple. The result directly addresses a conjecture in the literature and is likely to be of significant interest.
major comments (1)
- [§2, Lemma 2.5] Lemma 2.5 does not specify which Weierstrass model of E' is used to define the differential ω' and the period Ω_E'. The model (2.2) is not minimal, and if the period is taken from that model while the Tamagawa numbers are taken from the minimal model of Lemma 2.3, the Cassels formula is inconsistent. For the non-minimal model (2.2), the pullback satisfies φ*ω' = -ω, so the constant c defined by c·φ*ω' = ω would have |c| = 1, leading to |Sel_phihat| = 4, which would not force the rank bound. The proof should state explicitly that E' is taken with the minimal model of Lemma 2.3, that the isogeny is composed with the isomorphism to that model, and that for this model the pullback multiplier is -2 (so c = -1/2 and |c| = 1/2). With this clarification the period-ratio step is correct.
minor comments (5)
- [§2, Lemma 2.3] In the sentence 'the Kodaira symbols of E at 2, m, m+n and m+2n are equal to II, III*, I1 and I2', the letter E should be E'.
- [§2, Lemma 2.5] The equation 'c · φ*ω' = ω' is easy to misread; it would be clearer to write 'φ*ω' = (1/c) ω' or to state explicitly that c is the reciprocal of the pullback multiplier.
- [§2, Lemma 2.2] The computation of W_2(E) relies on Halberstadt's Table 1 and a specific list of invariants, but the table is not reproduced. This is acceptable, but a brief indication of how the table is read would improve readability.
- [§1] The remark that (1,b) is a point of infinite order on E_{b^2} for all but finitely many b is terse; one or two sentences explaining the use of Silverman's specialization theorem would be helpful.
- [§2] The paper does not give a numerical example (e.g., m=3,n=8) to illustrate the Selmer computation. Adding one would make the descent concrete and easier to follow.
Circularity Check
No circularity: the rank-zero proof is a self-contained 2-descent using external Cassels and Dokchitser-Dokchitser theorems; the only self-reference is an announced follow-up that plays no role in the proof.
full rationale
The derivation is self-contained and non-circular. The rank-zero conclusion is obtained by a direct 2-descent: Lemma 2.4 proves |Sel_phi|=2 by explicit local solubility on the curves C'_d, and Lemma 2.5 computes |Sel_phihat|=2 using the Cassels formula from [SS04, (6.2)] together with the period-ratio dichotomy |c| in {1, 1/2} from [DD15, Theorems 1.2 and 8.2], with the ratio forced to 1/2 by a parity contradiction against W(E)=1 from Lemma 2.2. These cited results are external theorems, not self-citations, and none of the numerical inputs is fitted to force the conclusion. The specialization t=(m+n)/(2m) is not tuned to the rank-zero target: the same descent works for every admissible triple of primes congruent to 3 mod 8. The only self-reference, the announced follow-up [Zyw25], is mentioned in the introduction and is not used anywhere in the proof. A reader's concern about model dependence in the period ratio is a correctness or applicability risk in importing [DD15], not circularity, because the manuscript invokes the theorem rather than redefining it; no step reduces to its own input.
Assumptions & free parameters
assumptions (6)
- standard math Green's theorem: every subset of primes with positive relative density contains infinitely many 3-term arithmetic progressions
- standard math Silverman's specialization theorem: rank(E_t(Q)) >= rank(E(Q(T))) for all but finitely many t
- standard math Tate's algorithm and local root number tables (Rohrlich 1993, Halberstadt 1998)
- standard math Cassels' formula for isogeny Selmer groups (Schaefer-Stoll equation (6.2), based on Cassels 1965)
- standard math Dokchitser-Dokchitser Theorems 1.2 and 8.2: period ratio under isogeny and parity of order of vanishing of L(E,s)
- standard math Schaefer-Stoll Lemma 6.1 exact sequence linking Sel_phi, Sel_2 and Sel_phihat
Cite this review
Pith. "Pith review of An elliptic surface with infinitely many fibers for which the rank does not jump." pith.science (2026). https://pith.science/paper/H6WVMQZC
@misc{pith2026250201026,
author = {Pith},
title = {Pith review of: An elliptic surface with infinitely many fibers for which the rank does not jump},
year = {2026},
howpublished = {\url{https://pith.science/paper/H6WVMQZC}},
note = {Machine review of arXiv:2502.01026}
}
abstract
Let $E$ be a nonisotrivial elliptic curve over $\mathbb{Q}(T)$ and denote the rank of the abelian group $E(\mathbb{Q}(T))$ by $r$. For all but finitely many $t\in \mathbb{Q}$, specialization will give an elliptic curve $E_t$ over $\mathbb{Q}$ for which the abelian group $E_t(\mathbb{Q})$ has rank at least $r$. Conjecturally, the set of $t\in\mathbb{Q}$ for which $E_t(\mathbb{Q})$ has rank exactly $r$ has positive density. We produce the first known example for which $E_t(\mathbb{Q})$ has rank $r$ for infinitely many $t\in\mathbb{Q}$. For our particular $E/\mathbb{Q}(T)$ which has rank $0$, we will make use of a theorem of Green on $3$-term arithmetic progressions in the primes to produce $t\in\mathbb{Q}$ for which $E_t$ has only a few bad primes that we understand well enough to perform a $2$-descent.
Forward citations
Cited by 2 Pith papers
-
Rank one elliptic curves and rank stability
There are infinitely many elliptic curves over any number field with rank 1, and for any quadratic extension L/K, infinitely many curves satisfy rank E(K) = rank E(L) = 1.
-
There are infinitely many elliptic curves over the rationals of rank 2
For m, m+16n^2, m+25n^2 primes congruent to 11 modulo 24, the curve y^2 = x^3 - 5(m+16n^2)x^2 + 4(m+16n^2)(m+25n^2)x has rank exactly 2, giving infinitely many rank-2 curves.
Reference graph
Works this paper leans on
-
[1]
[CP23] Jerson Caro and Hector Pasten, On the fibres of an elliptic surface where the rank does not jum p, Bull. Aust. Math. Soc. 108 (2023), no. 2, 276–282, DOI 10.1017/s0004972722001368. MR4 640089 ↑1, 1.1 [Cas65] J. W. S. Cassels, Arithmetic on curves of genus
-
[2]
On conjectures of Birc h and Swinnerton- Dyer, J
VIII. On conjectures of Birc h and Swinnerton- Dyer, J. Reine Angew. Math. 217 (1965), 180–199, DOI 10.1515/crll.1965.217.180. MR179169 ↑2 [CS82] J. W. S. Cassels and A. Schinzel, Selmer’s conjecture and families of elliptic curves , Bull. London Math. Soc. 14 (1982), no. 4, 345–348, DOI 10.1112/blms/14.4.345. MR663485 ↑1.2 [DD15] Tim Dokchitser and Vladi...
-
[1994]
MR1312368 ↑2, 2 [Sil83] , Heights and the specialization map for families of abelian v arieties, J. Reine Angew. Math. 342 (1983), 197–211, DOI 10.1515/crll.1983.342.197. MR703488 ↑1, 3 [SW04] William Stein and Mark Watkins, Modular parametrizations of Neumann-Setzer elliptic curv es, Int. Math. Res. Not. 27 (2004), 1395–1405, DOI 10.1155/S107379280413391...
Reviewed August 9, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.