arxiv: 2605.01623 · v1 · submitted 2026-05-02 · 🧮 math.AG · math.RA

Recognition: unknown

An algorithmic reduction to canonical forms for vector bundles on anisotropic conics

Diego Y\'epez, Eoin Mackall

Pith reviewed 2026-05-09 17:36 UTC · model grok-4.3

classification 🧮 math.AG math.RA

keywords vector bundlesanisotropic conicstransition matricescanonical formspolynomial algorithmindecomposable bundlesclutching covers

0 comments

The pith

A polynomial-time algorithm reduces transition matrices of vector bundles on real anisotropic conics to block-diagonal canonical forms.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper gives a polynomial complexity algorithm that takes transition matrices for vector bundles glued along a clutching cover of a real anisotropic conic and reduces them to canonical block-diagonal form. This supplies a constructive, efficient version of the classification that Grothendieck and Birkhoff obtained for the Riemann sphere. To make the algorithm work, the authors supply an elementary algebraic proof that every vector bundle on such a conic splits as a direct sum of indecomposable summands of rank at most two. The same methods are shown to apply verbatim to anisotropic conics over any base field. A reader would care because the result turns an abstract existence statement into an explicit computational procedure.

Core claim

We describe a polynomial complexity algorithm for reducing transition matrices, for vector bundles glued along a clutching-type cover of a real anisotropic conic, to canonical block diagonal forms. This is a generalization, to the real anisotropic form, of the classification of vector bundles on the Riemann sphere by their canonical diagonal forms due to Grothendieck and Birkhoff. We provide an elementary algebraic proof for the decomposition of vector bundles on real anisotropic conics into sums of indecomposable vector bundles of rank at most 2, and note that the methods extend immediately to anisotropic conics over arbitrary fields.

What carries the argument

The reduction algorithm that acts on transition matrices of a clutching-type cover and produces canonical block-diagonal forms by using the rank-at-most-two indecomposable decomposition.

If this is right

Bundles on these conics become classifiable by explicit canonical forms rather than by existence alone.
The classification procedure runs in polynomial time once a transition matrix is given.
The same reduction and decomposition hold verbatim for anisotropic conics over any field.
Cohomology groups and other invariants can be read off directly from the canonical block-diagonal representatives.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The algorithm could be coded directly in a computer algebra system to normalize explicit examples over the reals.
Similar reduction techniques might apply to vector bundles on other curves where indecomposable summands remain low-rank.
The focus on real coefficients raises the question of how the canonical forms detect real-algebraic features such as signature or positivity.

Load-bearing premise

Every vector bundle on the real anisotropic conic decomposes as a direct sum of indecomposable summands each of rank at most two.

What would settle it

A vector bundle on the real anisotropic conic whose transition matrix cannot be reduced by the algorithm to a block-diagonal form with blocks of size at most two, or whose reduction requires super-polynomial steps.

read the original abstract

We describe a polynomial complexity algorithm for reducing transition matrices, for vector bundles glued along a clutching-type cover of a real anisotropic conic, to canonical block diagonal forms. This is a generalization, to the real anisotropic form, of the classification of vector bundles on the Riemann sphere by their canonical diagonal forms due to Grothendieck and Birkhoff. To enable our algorithm, we provide an elementary algebraic proof for the result, due to Biswas-Nagaraj and Novakovic, of the decomposition of vector bundles on real anisotropic conics into sums of indecomposable vector bundles of rank at most 2. While our algorithm and our proof of this decomposition focus solely on the setting of a real anisotropic conic, our methods are immediately generalizable to anisotropic conics over arbitrary fields.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper turns the known decomposition of vector bundles on real anisotropic conics into an explicit polynomial-time algorithm for reducing transition matrices to canonical block-diagonal form.

read the letter

The paper gives a polynomial-time algorithm that reduces transition matrices for vector bundles on a real anisotropic conic to canonical block-diagonal form. It backs this with an elementary algebraic proof that every bundle decomposes into indecomposables of rank at most 2. The new part is the algorithm itself. The decomposition result was already known from Biswas-Nagaraj and Novakovic, but the authors supply a fresh elementary proof and then use it to construct the reduction procedure step by step. This turns an existence statement into something one can actually compute. The generalization note to other fields is a bonus if the details carry over cleanly. The manuscript does address the constructiveness issue. The proof proceeds by explicit algebraic manipulations rather than abstract existence arguments, and the algorithm follows by applying those manipulations to the transition matrix entries. Complexity stays polynomial because the operations involve matrices of size linear in the input degree. Soft spots are minor. The paper focuses tightly on the real anisotropic case, so the generalizability claim is stated but not fully expanded. No issues with circularity or undefined terms. This paper is for algebraic geometers who want computational tools for classifying bundles on real curves or conics. A reader interested in effective versions of classification theorems will find it useful. It deserves a serious referee because the claims are specific and checkable. I recommend sending it to peer review.

Referee Report

3 major / 2 minor

Summary. The paper claims to provide a polynomial-time algorithm that, given a transition matrix for a vector bundle on a real anisotropic conic (glued along a clutching cover), reduces it to a canonical block-diagonal form consisting of indecomposable summands of rank at most 2. This generalizes the Grothendieck-Birkhoff theorem from the Riemann sphere. An elementary algebraic proof is supplied for the decomposition theorem (previously due to Biswas-Nagaraj and Novakovic), and the methods are asserted to generalize immediately to anisotropic conics over arbitrary fields.

Significance. If the algorithm is polynomial and the decomposition proof is fully constructive, the result supplies an effective classification tool for vector bundles over real anisotropic conics, extending classical results to a setting without algebraic closure. The explicit linkage of an algebraic decomposition proof to a reduction procedure would be a notable contribution, particularly if it avoids non-constructive existence arguments and yields bounded complexity.

major comments (3)

[§3] §3 (Decomposition theorem): The algebraic proof of the rank-≤2 indecomposable decomposition proceeds by induction on rank and explicit matrix manipulations, but it is not shown that the inductive step produces the summands in time polynomial in the degree of the input transition matrix; the auxiliary linear systems solved at each step have degree that appears to grow with the input without an explicit bound.
[§4.1–4.3] §4.1–4.3 (Algorithm description): The reduction procedure is described via a sequence of elementary row and column operations on the clutching matrix, but no pseudocode, complexity analysis, or proof that each step terminates in O(d^k) operations for fixed k (where d is the degree of the input) is supplied; the claim of polynomial complexity therefore rests on an unverified assertion that the decomposition proof directly yields the algorithm.
[Theorem 5.1] Theorem 5.1 (generalization statement): The assertion that the methods are “immediately generalizable” to arbitrary fields is stated without any indication of the modifications required for the base field or for the clutching cover, making the scope of the result unclear.

minor comments (2)

[Introduction] The abstract and introduction cite Biswas-Nagaraj and Novakovic but do not clarify which parts of the decomposition are new versus reproduced; a short comparison paragraph would help.
[§2] Notation for the clutching cover and transition matrices is introduced without a dedicated preliminary section; readers unfamiliar with the real anisotropic conic setting must infer definitions from the examples.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the careful reading and for highlighting points where the manuscript would benefit from greater explicitness on complexity and scope. We address each major comment below and will revise the paper accordingly.

read point-by-point responses

Referee: [§3] §3 (Decomposition theorem): The algebraic proof of the rank-≤2 indecomposable decomposition proceeds by induction on rank and explicit matrix manipulations, but it is not shown that the inductive step produces the summands in time polynomial in the degree of the input transition matrix; the auxiliary linear systems solved at each step have degree that appears to grow with the input without an explicit bound.

Authors: We agree that an explicit polynomial bound on the degree of the auxiliary systems is missing. In the inductive step the input transition matrix has entries of degree at most d; each row/column operation and the subsequent linear algebra over the function field produces new polynomials whose degrees remain bounded by 2d (by the explicit form of the clutching functions on the anisotropic conic). We will insert a short lemma after the induction that records this degree bound and shows that each inductive step costs O(d^3) arithmetic operations, yielding an overall polynomial complexity for the decomposition. revision: yes
Referee: [§4.1–4.3] §4.1–4.3 (Algorithm description): The reduction procedure is described via a sequence of elementary row and column operations on the clutching matrix, but no pseudocode, complexity analysis, or proof that each step terminates in O(d^k) operations for fixed k (where d is the degree of the input) is supplied; the claim of polynomial complexity therefore rests on an unverified assertion that the decomposition proof directly yields the algorithm.

Authors: The referee is correct that the current text supplies neither pseudocode nor a self-contained complexity proof. We will add a new subsection 4.4 containing (i) pseudocode for the full reduction algorithm, (ii) a direct reference from each algorithmic step to the corresponding constructive step in the §3 proof, and (iii) a theorem stating that the total number of arithmetic operations is O(d^4). This will make the polynomial-complexity claim fully rigorous rather than implicit. revision: yes
Referee: [Theorem 5.1] Theorem 5.1 (generalization statement): The assertion that the methods are “immediately generalizable” to arbitrary fields is stated without any indication of the modifications required for the base field or for the clutching cover, making the scope of the result unclear.

Authors: We accept that the single sentence in the abstract and the brief remark at the end of §5 are insufficient. We will replace the claim with a short paragraph that explicitly lists the required changes: replace ℝ by an arbitrary field k, require the conic to be anisotropic over k, keep the same clutching cover (two copies of the affine line glued along the multiplicative group), and note that all matrix manipulations remain valid over k because they use only the ring structure of k[t] and the anisotropy hypothesis to guarantee the existence of the required splittings. No new ideas are needed, but the modifications must be spelled out. revision: yes

Circularity Check

0 steps flagged

No significant circularity: self-contained algebraic proof enables algorithm

full rationale

The paper supplies its own elementary algebraic proof of the decomposition of vector bundles into indecomposable summands of rank at most 2, rather than treating the Biswas-Nagaraj/Novakovic result as an unverified load-bearing citation. This proof is presented as original and is explicitly stated to enable the polynomial-complexity reduction algorithm for transition matrices. No equations, definitions, or steps in the abstract or described derivation chain reduce the claimed algorithm or decomposition to fitted parameters, self-referential quantities, or prior self-citations by construction. The generalization statement is ancillary and does not underpin the core claims. The derivation chain is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claims rest on the geometric properties of real anisotropic conics and on the existence of a clutching-type cover; the decomposition into rank ≤2 indecomposables is treated as a theorem to be proved rather than an axiom.

axioms (1)

domain assumption Vector bundles on a real anisotropic conic admit a decomposition into indecomposable summands of rank at most 2
Invoked as the foundation for the reduction algorithm; the paper supplies an elementary proof of this statement.

pith-pipeline@v0.9.0 · 5430 in / 1151 out tokens · 41841 ms · 2026-05-09T17:36:32.548340+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

22 extracted references · 9 canonical work pages

[1]

Transactions of the American Mathematical Society , volume=

Singular points of ordinary linear differential equations , author=. Transactions of the American Mathematical Society , volume=. 1909 , publisher=

1909
[2]

, author=

CLASSIFICATION OF REAL ALGEBRAIC VECTOR BUNDLES OVER THE REAL ANISOTROPIC CONIC. , author=. International Journal of Mathematics , volume=
[3]

Bulletin des sciences mathematiques , volume=

Absolutely split real algebraic vector bundles over a real form of projective space , author=. Bulletin des sciences mathematiques , volume=. 2007 , publisher=

2007
[4]

Journal of the Australian Mathematical Society , volume=

Vector bundles over a nondegenerate conic , author=. Journal of the Australian Mathematical Society , volume=. 2009 , publisher=

2009
[5]

Theorie der algebraischen Functionen einer Ver

Weber, H and Dedekind, R , journal=. Theorie der algebraischen Functionen einer Ver
[6]

Sur la classification des fibr

Grothendieck, Alexander , journal=. Sur la classification des fibr. 1957 , publisher=

1957
[7]

, TITLE =

Hazewinkel, Michiel and Martin, Clyde F. , TITLE =. J. Pure Appl. Algebra , FJOURNAL =. 1982 , NUMBER =. doi:10.1016/0022-4049(82)90037-8 , URL =

work page doi:10.1016/0022-4049(82)90037-8 1982
[8]

Bulletin des Sciences Math

Absolutely split locally free sheaves on Brauer--Severi varieties of index two , author=. Bulletin des Sciences Math. 2012 , publisher=

2012
[9]

Bulletin des Sciences Math

Absolutely split locally free sheaves on proper k-schemes and Brauer--Severi varieties , author=. Bulletin des Sciences Math. 2024 , publisher=

2024
[10]

1994 , PAGES =

Rosenberg, Jonathan , TITLE =. 1994 , PAGES =. doi:10.1007/978-1-4612-4314-4 , URL =

work page doi:10.1007/978-1-4612-4314-4 1994
[11]

Chernousov, Vladimir and Merkurjev, Alexander , TITLE =. Compos. Math. , FJOURNAL =. 2006 , NUMBER =. doi:10.1112/S0010437X06002363 , URL =

work page doi:10.1112/s0010437x06002363 2006
[12]

Ma, Qixiao , TITLE =. J. Algebra , FJOURNAL =. 2022 , PAGES =. doi:10.1016/j.jalgebra.2021.08.029 , URL =

work page doi:10.1016/j.jalgebra.2021.08.029 2022
[13]

2017 , PAGES =

Gille, Philippe and Szamuely, Tam\'as , TITLE =. 2017 , PAGES =

2017
[14]

Brauer groups in ring theory and algebraic geometry (

Artin, Michael , TITLE =. Brauer groups in ring theory and algebraic geometry (. 1982 , ISBN =

1982
[15]

Journal of Mathematical Sciences , year =

Ephremidze, Lasha , title =. Journal of Mathematical Sciences , year =. doi:10.1007/s10958-025-07685-5 , url =

work page doi:10.1007/s10958-025-07685-5
[16]

Linear Algebra Appl

Storjohann, Arne and Labahn, George , TITLE =. Linear Algebra Appl. , FJOURNAL =. 1997 , PAGES =. doi:10.1016/0024-3795(95)00743-1 , URL =

work page doi:10.1016/0024-3795(95)00743-1 1997
[17]

Wilkening, Jon and Yu, Jia , TITLE =. J. Symbolic Comput. , FJOURNAL =. 2011 , NUMBER =. doi:10.1016/j.jsc.2010.06.025 , URL =

work page doi:10.1016/j.jsc.2010.06.025 2011
[18]

Villard, Gilles , TITLE =. J. Symbolic Comput. , FJOURNAL =. 1995 , NUMBER =. doi:10.1006/jsco.1995.1050 , URL =

work page doi:10.1006/jsco.1995.1050 1995
[19]

Computing the Roots of Twisting Sheaves over the Projective Line Arising from Monodromy Representations , author =. Ann. Sc. Norm. Super. Pisa Cl. Sci. , year =
[20]

arXiv preprint arXiv:2507.07080 , year=

Monodromy Representations: Decomposing Rank-Three Bundles over the Projective Line with Three Marked Points , author=. arXiv preprint arXiv:2507.07080 , year=

work page arXiv
[21]

Publications Math\'

Nicholas Katz , title =. Publications Math\'. 1970 , pages =

1970
[22]

Bulletin of the American Mathematical Society , volume=

Exponential sums over finite fields and differential equations over the complex numbers: some interactions , author=. Bulletin of the American Mathematical Society , volume=