REVIEW 3 major objections 3 minor
Local constancy of reduction type and related invariants for curves in $p$-adic families
T0 review · 3 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read Small p-adic changes leave a curve's reduction type fixed.
desk verdict The abstract's Galois-representation claim is false on the standard reading; the rest of the paper is unverifiable from the abstract alone. 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 central object is a family of curves over a discretely valued field, defined as the set of curves obtained by perturbing the coefficients of the defining equations. The mechanism carrying the argument is the valuation topology on the coefficient space: the paper shows that under this topology the reduction type, and with it the Tamagawa number, the BSD fudge factor, and the Galois representation, are locally constant, meaning each invariant takes a single value on a sufficiently small open ball around any coefficient vector.
What would settle it
One could compute the reduction type and Tamagawa number for a sequence of hyperelliptic curves over $\mathbb{Q}_p$ whose coefficient vectors converge $p$-adically to a limit; if any invariant changes infinitely often along the sequence, the local constancy claim is false.
Extended reading notes
Core claim
The central discovery is that reduction type, Tamagawa number, the Birch and Swinnerton-Dyer fudge factor, and the Galois representation are all locally constant functions on the coefficient space of a family of curves defined over a discretely valued field, where the family is taken to be a set of curves obtained by perturbing the defining equations. The author states this as a theorem: within the topology induced by the valuation, every curve has a neighborhood on which these invariants do not change. The reduction type is the key invariant, and the other invariants are shown to share the same local constancy.
Load-bearing premise
The result rests on the assumption that for every sufficiently small perturbation, all the curves in the family admit one common proper model over the base ring, so that a single special fiber exists from which the reduction invariants can be read; if that common model does not exist, local constancy could fail.
Editorial extensions
If this is right
- If the result is correct, the reduction type of a curve over a discretely valued field is an invariant of its $p$-adic neighborhood, so any two curves whose defining equations are sufficiently close in the valuation topology share the same reduction type.
- The Tamagawa number, which records the component structure of the special fiber, is locally constant, so arithmetic formulas that sum Tamagawa numbers over primes are stable under small coefficient changes.
- The Birch and Swinnerton-Dyer fudge factor is locally constant, meaning the local contribution to the BSD conjecture remains unchanged under small $p$-adic perturbations of the curve.
- The Galois representation attached to a curve is locally constant, so the action of the absolute Galois group on its étale cohomology is unchanged on a $p$-adic neighborhood of the coefficients.
Reading between the lines
- A testable extension would be to compute, for explicit families of hyperelliptic curves over $\mathbb{Q}_p$, the radius of the largest ball on which the reduction type is constant; the paper does not give such radii.
- If local constancy holds for curves, the same perturbation argument may apply to other arithmetic objects such as abelian varieties or higher-dimensional varieties, where reduction-type invariants might also be locally constant; this is an inference, not a claim of the paper.
- The result suggests an algorithmic shortcut: instead of computing reduction invariants for a continuum of curves, one can compute them at a single representative coefficient vector in each sufficiently small ball.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript, as submitted, consists of an abstract alone. It claims that for a family of curves over a discretely valued field obtained by perturbing the coefficients of defining equations, the reduction type is locally constant in the valuation topology, and that analogous local constancy holds for the Tamagawa number, the Birch and Swinnerton-Dyer fudge factor, and the Galois representation. The document contains no definitions, hypotheses, theorem statements, or proofs.
Significance. If the intended statement were made precise and proved, the result would be a useful complement to the theory of reduction in p-adic families, with potential applications to the arithmetic of curves in families. However, the submitted text provides no evidence for these claims, and the Galois representation assertion is ambiguous and, on its most natural reading, contradicted by standard examples. The contribution cannot be evaluated from the present text.
major comments (3)
- [Abstract] The assertion that 'the Galois representation' is locally constant is ambiguous and, under the usual meaning (the rational p-adic or ℓ-adic étale cohomology representation), false. For example, the elliptic curves E_n: y^2 = x^3 + 7^n x + 1 over Q_7 all have good ordinary reduction modulo 7 (they reduce to the same curve y^2 = x^3 + 1), yet their p-adic Tate modules V_7(E_n) are not generally isomorphic because their filtered (φ,N)-modules have extension data that depends on the j-invariant, which varies in every 7-adic neighbourhood of the family. A local constancy theorem cannot hold for the full representation; the manuscript must state precisely which invariant (e.g., the semisimplified mod-p representation or the Newton polygon) is claimed to be locally constant, and it must prove that weaker statement.
- [Abstract] The paper defines a 'family' only as 'a set of curves obtained by perturbing the coefficients of the defining equations'. This does not specify the parameter space, the base ring, the choice of model over the base, or the smoothness/properness hypotheses. In particular, the abstract does not say that the family avoids the discriminant locus or that a simultaneous regular (or semistable) model exists. Without such hypotheses, local constancy of the reduction type is not guaranteed. The full statement must include a precise base scheme and a proper flat model over it whose generic fiber is the family.
- [Entire manuscript] The submission contains no body: there are no formal theorem statements, no proofs, no definitions of 'reduction type' or the 'fudge factor'. The abstract's promises ('we will show', 'we also derive') are not backed by verifiable mathematics. This is not a refereable paper in its current state; the authors need to submit a complete manuscript with precise statements and full arguments.
minor comments (3)
- [Abstract] The term 'Birch and Swinnerton-Dyer fudge factor' is informal and should be defined; presumably it denotes the leading coefficient of the L-function at s=1 (the product of period, Tamagawa number, regulator, and torsion factors), but the paper should say so explicitly.
- [Abstract] The phrase 'topology induced by the valuation' should be clarified: for each coefficient vector, the paper should indicate whether the neighborhood is a p-adic ball in the coefficient space and how the curves are parametrized by the coefficients.
- [References] The abstract gives no references to prior results on constancy of reduction or Néron models in p-adic families; for instance, the constancy of the Euler characteristic and the existence of Néron models in proper flat families (e.g., Deligne's work) should be cited to frame the claimed result.
Circularity Check
No circularity identified from the available text.
full rationale
The submission contains only the abstract; the full derivation chain is not present, so there is no equation, fitted parameter, or self-citation to compare. The abstract's definition of 'family' as 'a set of curves obtained by perturbing the coefficients of the defining equations' does not, by itself, define reduction type, the Tamagawa number, the BSD fudge factor, or the Galois representation in terms of the conclusion of local constancy. Local constancy of these invariants under p-adically small coefficient perturbations is a substantive geometric claim that would need a proper flat model argument over a base ring; the abstract does not reduce it to an identity. The skeptic's objection that the natural reading of 'the Galois representation' is false for p-adic Tate modules is a correctness concern, not a circularity concern, and cannot be resolved without the proof. No visible step reduces to its own input, and there are no self-citations to evaluate. Hence the appropriate finding is no significant circularity.
Assumptions & free parameters
assumptions (3)
- standard math The standard theory of reduction types, regular and semistable models, Tamagawa numbers, and Galois representations of curves over discretely valued fields is taken as background.
- domain assumption A family of curves may be represented as coefficient perturbations of defining equations, and every such perturbation yields a curve of the same kind over the fraction field of the discretely valued field.
- domain assumption The invariants in question are well-defined and finite on an open neighborhood of the coefficient space, so that local constancy statements have a meaningful domain.
Cite this review
Pith. "Pith review of Local constancy of reduction type and related invariants for curves in $p$-adic families." pith.science (2026). https://pith.science/paper/EDYV4FWG
@misc{pith2026250812329,
author = {Pith},
title = {Pith review of: Local constancy of reduction type and related invariants for curves in $p$-adic families},
year = {2026},
howpublished = {\url{https://pith.science/paper/EDYV4FWG}},
note = {Machine review of arXiv:2508.12329}
}
read the original abstract
We investigate the behaviour of the reduction type and related invariants of curves in families of curves over a discretely valued field. By a family, we will mean a set of curves obtained by perturbing the coefficients of the defining equations. We will show that the reduction type in these families is locally constant in the topology induced by the valuation. We also derive local constancy results for some related invariants, such as the Tamagawa number, the Birch and Swinnerton-Dyer 'fudge factor' and the Galois representation.
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.