REVIEW 2 major objections 3 minor 15 references
Birational equivalence of Severi-Brauer varieties
T0 review · 2 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read Amitsur's conjecture for Severi–Brauer varieties is proved when dimension plus one has at least two distinct prime factors.
desk verdict Important theorem, elegant method, but the proof of Theorem 2 has an unaddressed separability gap that needs a fix before the paper is complete. 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 load-bearing object is the tensor (Brauer) product P1 ⊗ P2 of Severi–Brauer varieties — the unique Severi–Brauer variety containing P1 × P2 as a Segre-embedded subvariety. The paper shows this twisted product is birationally an ordinary product times a projective space when the dimensions are coprime. The engine is Proposition 17, a Weil restriction formula: for a Severi–Brauer variety P of dimension n and a separable extension K/k of degree m + 1 with gcd(m + 1, n + 1) = 1, the Weil restriction ℜ_{K/k}(P_K) — the variety whose k-points are K-points of P — is birational to P × $P^{{mn}}$. The proof passes through the Grassmannian Grass(m, P) of m-dimensional linear subspaces of P and its universal pointed family; restricting canonical classes shows the universal family is birationally a product, and the formula follows.
What would settle it
Compute the Weil restriction in the smallest nontrivial coprime case: a Severi–Brauer surface P (n = 2) over a field k and a separable quadratic extension K/k (m + 1 = 2), and check whether ℜ_{K/k}(P_K) is birationally equivalent to P × $P^{2}$. Since equality over the algebraic closure is already known, the failure could only manifest through twisted forms; one explicit way is to compare the canonical ring or Chow groups of the two fourfolds. Any example where they differ disproves Proposition 17 and the main theorems.
Extended reading notes
Core claim
The central theorem (Theorem 1) states: let P and Q be Severi–Brauer varieties of the same dimension over a field k, with dim P + 1 having at least two distinct prime factors. Then P and Q generate the same subgroup of the Brauer group Br(k) if and only if P is birationally equivalent to Q. Amitsur proved the forward direction; the paper proves the converse. The proof is a consequence of Theorem 2: for Severi–Brauer varieties P1, P2 with (dim P1 + 1, dim P2 + 1) = 1, there is a birational equivalence P1 ⊗ P2 bir ∼ P1 × P2 × $P^{{r}}$ with r = dim P1 · dim P2. This converts a twisted product into an ordinary product plus a projective space, allowing the author to swap factors one by one after a primary decomposition. The whole argument reduces to the new Weil restriction formula (17.1) for a Severi–Brauer variety base-changed to a separable extension of coprime degree.
Load-bearing premise
The proof rests on the new Weil restriction formula (17.1): when a Severi–Brauer variety of dimension n is base-changed to a separable extension of degree m + 1 coprime to n + 1, its Weil restriction is birational to the original variety times a projective space of dimension mn; the paper gives only a half-page sketch of this formula, and a gap there would invalidate Theorem 2 and hence Theorem 1.
Editorial extensions
If this is right
- Amitsur's conjecture is settled for every Severi–Brauer variety whose index is not a prime power, since such an index forces dim P + 1 to have at least two distinct prime factors.
- The birational product formula (2.1) gives an explicit way to split off a projective space from a tensor product of coprime-degree varieties, reducing questions about twisted products to ordinary products.
- In the proof of Theorem 1, the non-minimal case is handled without the two-prime-factor assumption, so the only outstanding cases are minimal varieties whose dimension plus one is a prime power.
- Repeated application of Theorem 2 yields new instances of the product conjecture for arbitrary products of Severi–Brauer varieties (Conjecture 4), and the paper shows (4.1) implies (4.2) unconditionally.
- The failure of Proposition 17 without coprimality (Remark 17.5) explains why the prime-power case is genuinely different: Weil restriction can then detect whether P has a K-point.
Reading between the lines
- A natural extension is to test whether the Weil restriction formula (17.1) holds for generalized Severi–Brauer varieties parametrizing rank-r subspaces of a division algebra; the Grassmannian bundle argument suggests an analogous product formula when the relevant degrees are coprime.
- If Proposition 17 could be strengthened to non-coprime degrees up to a correction term depending on the splitting behavior of K, Theorem 2 would likely extend to all dimensions, completing Amitsur's conjecture.
- The canonical-class step in the proof of (17.1) indicates that the right invariant controlling these birational cancellations is the pair of integers (m + 1, n + 1) and their gcd; one could mine the formula for more refined birational invariants of Weil restrictions.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves Amitsur's conjecture for Severi-Brauer varieties whose index is not a prime power. The main theorem (Theorem 1) states that two Severi-Brauer varieties of the same dimension with dim(P)+1 having at least two distinct prime factors are birationally equivalent if and only if they generate the same subgroup of the Brauer group. The argument reduces to a new birational product formula (Theorem 2) for tensor products of Severi-Brauer varieties of coprime dimensions, and the key new ingredient is a Weil restriction formula (Proposition 17) for separable field extensions.
Significance. If correct, this would be a significant advance on a long-standing open problem in the theory of Severi-Brauer varieties. The paper is clearly written, builds on work of Amitsur, Roquette, and Krashen, and introduces a potentially useful Weil restriction method. However, the proof of Theorem 2 contains a serious gap concerning the separability of the splitting field used in Section 18, so the main theorem is not established as written.
major comments (2)
- [Section 18, equations (18.1)-(18.2)] The proof of Theorem 2 chooses a splitting field K/k of P2 of degree dim(P2)+1, but Proposition 17 and Definition 15 require K/k to be separable. Over imperfect fields, a Severi-Brauer variety of prime-power index need not have a separable splitting field of degree equal to its index; Albert's non-separable p-algebras provide examples. For such P2, the Weil restriction ℜ_{K/k}(L) used in (18.1) is not defined by the paper's Definition 15, and the birational formula (18.2) from Proposition 17 cannot be invoked. Since Theorem 1 relies on Theorem 2 through the reduction in Section 13, this gap affects the central claim of the paper. The authors need either to prove an analogous statement for inseparable extensions or to provide a separate argument covering that case.
- [Proposition 17, equations (17.2)-(17.3)] The proof of Proposition 17, which is the only new ingredient of the paper, is very terse. In particular, the step from (17.2) to (17.3) asserts that the line bundle O_F(1) on each fiber F exists and is a line bundle on the total space, implying that the P^m-bundle is birationally trivial. This is a nontrivial claim about the Brauer class of the relative Severi-Brauer bundle and should be proved in full detail, since Proposition 17 is load-bearing for Theorem 2.
minor comments (3)
- [Definition 15] The restriction to separable extensions in the definition of Weil restriction should be explicitly recalled when K is chosen in Section 18, as the current proof does not mention this condition.
- [Section 13] In the proof of Theorem 1, the converse implication relies on Theorem 2, which is proven later in Section 18. This logical dependence is acceptable, but it would help the reader if the exposition noted explicitly that the proof of Theorem 1 assumes Theorem 2.
- [Abstract and Theorem 1] The abstract states the result for varieties whose index is not a prime power, while Theorem 1 assumes that dim(P)+1 has at least two distinct prime factors. These conditions differ when P is not minimal, and the discrepancy should be clarified for the reader.
Circularity Check
No significant circularity: the proof is self-contained and the only self-citations are background, not load-bearing.
full rationale
The paper's derivation chain is not circular. Theorem 1 is deduced from Theorem 2, and Theorem 2 is proved from Corollary 16 and Proposition 17. Proposition 17 is explicitly flagged as the only new result and is proved using Grassmannian bundle computations (17.2)-(17.4), the Weil-restriction formula (15.2), and a canonical-class/Brauer-class argument; none of these steps assumes the conclusion. The proof of Theorem 2 then constructs the relative Weil restriction of a twisted linear subvariety and uses Corollary 10 to control the Brauer class; the final birational equivalence is obtained by combining independent statements, not by re-using the theorem being proved. Theorem 1's proof is a straightforward algebra of Brauer classes plus Theorem 2; the subgroup condition is an input, not a consequence of birationality. The only self-citations are [Kol25] for background on Severi-Brauer varieties and [Kol05] for low-dimensional cases; neither is the load-bearing step, which rests on Amitsur, Roquette, Krashen, and Tregub. The paper's own note that 'all the steps can be found in [Roq64, Kra08]' is an attribution of method, not a circular appeal. The possible gap in Theorem 2's proof over imperfect fields, namely choosing a splitting field K of degree dim P2 + 1 without proving separability although Proposition 17 requires it, is a correctness or hypotheses issue, not a circularity: no equation or conclusion is defined in terms of the target result. Therefore no circular step is present.
Assumptions & free parameters
assumptions (5)
- standard math Standard properties of the Brauer group of a field, including that it is torsion and that the index divides the period and has the same prime factors.
- standard math For a central simple algebra of index n over k, there exists a finite separable splitting field of degree n (a maximal subfield).
- standard math Amitsur's theorem that birational equivalence of Severi-Brauer varieties implies equality of the generated subgroups of the Brauer group.
- standard math Roquette's inductive method and Krashen's product theorem for Severi-Brauer varieties.
- standard math Properties of Weil restriction, including that the Weil restriction of a projective space of dimension n over a degree m extension is birational to a projective space of dimension n times m.
Cite this review
Pith. "Pith review of Birational equivalence of Severi-Brauer varieties." pith.science (2026). https://pith.science/paper/T4QNMCBA
@misc{pith2026250524720,
author = {Pith},
title = {Pith review of: Birational equivalence of Severi-Brauer varieties},
year = {2026},
howpublished = {\url{https://pith.science/paper/T4QNMCBA}},
note = {Machine review of arXiv:2505.24720}
}
read the original abstract
We prove Amitsur's conjecture for Severi-Brauer varieties whose index is not a prime power.
Reference graph
Works this paper leans on
-
[1]
Adrian Albert, Structure of A lgebras , American Mathematical Society Colloquium Publications, vol
A. Adrian Albert, Structure of A lgebras , American Mathematical Society Colloquium Publications, vol. Vol. 24, American Mathematical Society, New York, 1939. 595
work page 1939
-
[2]
S. A. Amitsur, Generic splitting fields of central simple algebras, Ann. of Math. (2) 62 (1955), 8--43. 70624
work page 1955
-
[3]
21, Springer-Verlag, Berlin, 1990
Siegfried Bosch, Werner L \"u tkebohmert, and Michel Raynaud, N\'eron models, Ergebnisse der Mathematik und ihrer Grenzgebiete (3), vol. 21, Springer-Verlag, Berlin, 1990. 1045822 (91i:14034)
1990
-
[4]
101, Cambridge University Press, Cambridge, 2006
Philippe Gille and Tam \'a s Sz amuely, Central simple algebras and G alois cohomology , Cambridge Studies in Advanced Mathematics, vol. 101, Cambridge University Press, Cambridge, 2006. 2266528
work page 2006
-
[5]
I. N. Herstein, Noncommutative rings, The Carus Mathematical Monographs, vol. No. 15, Mathematical Association of America, ; distributed by John Wiley & Sons, New York, 1968. 227205
work page 1968
-
[6]
Amit Hogadi, Products of B rauer- S everi surfaces , Proc. Amer. Math. Soc. 137 (2009), no. 1, 45--50. 2439423
work page 2009
-
[7]
J \'a nos Koll \'a r, Conics in the G rothendieck ring , Adv. Math. 198 (2005), no. 1, 27--35. 2183248 (2006k:14064)
work page 2005
-
[8]
, Severi-Brauer varieties; a geometric treatment , https://arxiv.org/abs/1606.04368, 2025
arXiv 2025
Show all 15 references
-
[9]
Daniel Krashen, Severi- B rauer varieties of semidirect product algebras , Doc. Math. 8 (2003), 527--546. 2029172
2003
-
[10]
Pure Appl
, Birational maps between generalized S everi- B rauer varieties , J. Pure Appl. Algebra 212 (2008), no. 4, 689--703. 2363485
2008
-
[11]
Reine Angew
Peter Roquette, Isomorphisms of generic splitting fields of simple algebras, J. Reine Angew. Math. 214/215 (1964), 207--226. 166215
1964
-
[12]
VIII, Hermann, Paris, 1962, Actualit\'es Scientifiques et Industrielles, No
Jean-Pierre Serre, Corps locaux, Publications de l'Institut de Math\'ematique de l'Universit\'e de Nancago, vol. VIII, Hermann, Paris, 1962, Actualit\'es Scientifiques et Industrielles, No. 1296. 150130
1962
-
[13]
67, Springer-Verlag, New York-Berlin, 1979, Translated from the French by Marvin Jay Greenberg
, Local fields, Graduate Texts in Mathematics, vol. 67, Springer-Verlag, New York-Berlin, 1979, Translated from the French by Marvin Jay Greenberg. 554237
1979
-
[14]
S. L. Tregub, Birational equivalence of B rauer- S everi manifolds , Uspekhi Mat. Nauk 46 (1991), no. 6(282), 217--218. 1164209
1991
-
[15]
23, Birkh\"auser, Boston, MA, 1982, With appendices by M
Andr\'e Weil, Adeles and algebraic groups, Progress in Mathematics, vol. 23, Birkh\"auser, Boston, MA, 1982, With appendices by M. Demazure and Takashi Ono. 670072
1982
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.