REVIEW
On the $abc$ conjecture
Algebraic Geometry · Submitted 3 Nov 2026 · 28 references
Pith's verdict: A serious near-completion of the Parshin–Moret-Bailly program whose intersection machinery mostly checks out, but §6.3.3 conflates the actual fiber with an auxiliary quadratic twist, leaving Theorem 1 unproven as written.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
The paper proves effective Mordell for genus-2 curves and, via a Kodaira fibration argument, derives the weak abc bound c ≤ A(ε)·N^(24+ε).
The reading
What carries the argument
The engine is a two-sided estimate of a single arithmetic intersection number. The paper constructs the Kodaira-Parshin fibration $\phi: X \to Y$, a non-isotrivial smooth family of genus-6 curves over a finite étale cover $Y$ of the genus-2 curve $C$, with two sections whose images are $B_1$ and $B_2$, and studies the line bundle $L = \omega_{X/Y} - B$ with $B = 2B_1 + 2B_2$. On the fiber over a lifted point $P'$ it restricts to $\rho'^*(\omega_1 - D)$, the pullback of a degree-one line bundle under the double cover defined by a square root $D$ of $P'_1 + P'_2$, and the intersection classes on $X$ obey $L^2 = -\tfrac{5}{3}\,\omega^2$ (numerically $-20k$ and $12k$). One bound on $L'^2_{P'}$ comes from Minkowski's theorem on successive minima applied to the two-dimensional space $H = H^0(L'_{P'})$, after Proposition 7's alteration argument supplies an integral model; the opposing bound comes from the degree estimate $\deg \phi_*(L) \ge -3k$ together with Corollary 1, which converts a degree-zero adelic class into an inequality up to an ample height. Proposition 3 — the admissible-metrics inequality $-\overline{L_1}\cdot\overline{L_2} \le d_2 h(s_1) + d_1 h(s_2)$ — turns the two sections $s_1, s_2$ into an intersection bound, and Proposition 6 is the base-point-free step guaranteeing that $s_1, s_2$ have no common zero, which is where the square-root divisor $D$ must be chosen.
Extended reading notes
Core claim
The load-bearing result is Theorem 1: for the non-isotrivial smooth fibration $\phi: X \to Y$ whose fibers are the genus-6 curves built from a genus-2 curve by the Kodaira construction, every rational point $P \in Y(F)$ satisfies $\omega_P^2 \le (36+\varepsilon)\,\delta(F) + 0(d)$, where $\delta(F) = \log|\Delta_F|$ and $d = [F:\mathbb{Q}]$. Because the Deligne pairing makes $\omega_P^2$ a logarithmic height on $Y$ and the cover $Y \to C$ is finite étale, this transfers to the base curve: $h(P) \le (12+\varepsilon)\,\delta(F)/[F:\mathbb{Q}] + \beta(\varepsilon)$, the effective Mordell bound that the introduction calls Theorem A and Section 7 proves as Theorem 2. Following the template of [16], the same inequality yields a weak Szpiro bound $\log|\Delta(E)| \le (144(1+\log 24)+\varepsilon)\log N(E) + \beta_2(\varepsilon)$ for semistable elliptic curves over $\mathbb{Q}$, a uniform weak $abc$ theorem over number fields, $\log H(a,b,c) \le (24+\varepsilon)\log N(a,b,c) + (30+\varepsilon)\log|\Delta_K|/[K:\mathbb{Q}] + \beta_3(\varepsilon)$, and, via the argument of [8], the absence of Siegel zeros for quadratic characters of negative discriminant.
Load-bearing premise
The chain that produces the bound collapses if the degree-one divisor $D$ chosen in §6.3.3 — with $2D = P'_1 + P'_2$ and $D + R_i$ outside the $\theta$ divisor for $i=1,2$ — is not the same square root as the one defining the actual fiber $X_{P'}$ of the Kodaira fibration: the paper asserts the equality $\rho'^*(\omega_1 - D) = L'_{P'}$ without deriving it, and without that equality the sections of $L'_{P'}$ may have a common zero, so inequality (34) does not follow.
Editorial extensions
If this is right
- Rational points on any genus-2 curve over a number field are bounded in height by $(12+\varepsilon)\,\delta(F)/[F:\mathbb{Q}] + \beta(\varepsilon)$, so for each fixed $F$ the set $C(F)$ lies in a bounded, effectively reachable region — an effective Mordell statement.
- Coprime solutions of $a+b=c$ satisfy $c \le A(\varepsilon)\,N^{24+\varepsilon}$, and the uniform field version gives $\log H(a,b,c) \le (24+\varepsilon)\log N(a,b,c) + (30+\varepsilon)\log|\Delta_K|/[K:\mathbb{Q}] + \beta_3(\varepsilon)$.
- Semistable elliptic curves over $\mathbb{Q}$ satisfy the weak Szpiro bound $\log|\Delta(E)| \le (144(1+\log 24)+\varepsilon)\log N(E) + \beta_2(\varepsilon)$.
- For every fundamental discriminant $-d < 0$, $L(s,\chi_d)$ has no real zero in the interval $1 - c/\log d < s \le 1$.
- Because the constants are explicit and uniform in the field, the consequences of $abc$ that follow through uniform $abc$-type inequalities come out effective, with weaker exponents than the conjectural ones.
Reading between the lines
- Editorial: the $abc$ exponent in Theorem 4 is structurally $\tfrac{2}{3}$ of the constant in Theorem 1, via the factors $\tfrac{1}{3}$ from (40)–(42), $\tfrac{5}{2}$ from the degree of $\pi^*\mathcal{O}(1)$ on $y^2+y=x^5$, and $\tfrac{4}{5}$ from the discriminant bound (51). Any improvement in the constant $36$ of Theorem 1 would lower the $abc$ exponent proportionally; the paper does not claim su
- Editorial: the identification step in §6.3.3 — that the chosen square-root divisor $D$ is the twist defining the actual fiber — is a local question about the restriction of the Kodaira line bundle, decoupled from the height machinery; if it fails, the repair has to come from constructing $L$ with a prescribed fiber restriction rather than from any constant in the Minkowski step.
- Editorial: the bound is concrete enough to test numerically on the curve $y^2 + y = x^5$: enumerating all points over small quadratic fields and comparing their canonical heights against $(12+\varepsilon)\,\delta(F)/[F:\mathbb{Q}]$ would pin down the effective constant $\beta(\varepsilon)$ and probe the uniformity in $F$ that the theorem asserts.
Original abstract
Using adelic intersection theory and following the method envisioned by Parshin and Moret-Bailly, we prove a conjecture of Vojta for curves of genus two and a weak form of the $abc$ conjecture
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Using adelic intersection theory and following the method envisioned by Parshin and Moret-Bailly, we prove a conjecture of Vojta for curves of genus two and a weak form of the $abc$ conjecture
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Using adelic intersection theory and following the method envisioned by Parshin and Moret-Bailly, we prove a conjecture of Vojta for curves of genus two and a weak form of the $abc$ conjecture | contribution 0.65 | 0.96 | 0.95 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.