REVIEW 3 major objections 5 minor 24 references
On the injective self-maps of algebraic varieties
T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read A self-map of an algebraic variety that is injective on the complement of a codimension-at-least-2 closed subset is an automorphism if the variety is a surface, non-singular, or locally a complete intersection regular in codimension 2.
desk verdict New cases of the Miyanishi conjecture with clean smooth/surface proofs, but the l.c.i./Q-factorial claims depend on an unverified hypothesis in Kallström's corollary — worth refereeing, likely fixable. 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 key mechanism is the sheaf of Kähler differentials Ω_X, the algebraic analogue of cotangent vectors, together with the canonical map φ*(Ω_X) → Ω_X → Ω_{X/X} → 0. Proving that the first map is an isomorphism forces Ω_{X/X}=0, i.e. φ is unramified, hence quasi-finite. Because both sheaves are reflexive in the relevant classes, an isomorphism on X\Y, where codim Y ≥ 2, extends globally via Proposition 2.9. In the locally complete intersection and Q-factorial settings, a criterion of Kallström ([14, Corollary 4.9]) is invoked to turn submersiveness, the vanishing of the critical module of the differential dφ, into étaleness for the smooth-locus restriction. Lemma 2.7 then converts quasi-finiteness plus birationality into an automorphism by Zariski's Main Theorem and Ax's theorem.
What would settle it
Produce a normal locally complete intersection variety regular in codimension 2 over C, with an endomorphism φ injective on X\Y for some closed Y of codimension at least 2, and with φ not quasi-finite, for instance a point at which the differential map φ*(Ω_X) → Ω_X is not surjective; such an example would contradict Theorem 1.2(iii) and show Kallström's criterion was misapplied.
Extended reading notes
Core claim
The paper's central claim is that the Miyanishi conjecture holds for normal algebraic varieties in three classes: algebraic surfaces, non-singular varieties, and locally complete intersection varieties satisfying Serre's condition (R2), meaning they are regular in codimension 2. In each case, an endomorphism φ that is injective on X\Y with codim_X Y ≥ 2 is shown to be quasi-finite; once quasi-finite, birationality plus Zariski's Main Theorem makes φ an open immersion, and Ax's theorem upgrades injectivity to an automorphism. The quasi-finiteness is obtained by showing the Kähler differential map φ*(Ω_X) → Ω_X is an isomorphism, so φ is unramified, using reflexivity results to extend the isomorphism from the open set X\Y across the codimension-2 set Y. A related theorem shows the restriction of φ to the smooth locus is an automorphism under hypotheses such as surjectivity with c=2, Q-factoriality, local completeness of intersection, or exceptional codimension at least 3; for varieties with finitely many singular points, this smooth-locus automorphism is enough to conclude φ itself is an automorphism.
Load-bearing premise
The locally complete intersection and Q-factorial parts depend on a cited purity criterion whose hypotheses are verified through references rather than proved in the stated generality, so if that criterion does not apply at the full stated generality, those cases fail.
Editorial extensions
If this is right
- In the three classes covered, Miyanishi's conjecture is settled: no non-automorphism can be injective off a codimension-2 closed subset.
- Under one of Theorem 1.3's conditions, φ maps the smooth locus to itself and gives an automorphism of the smooth locus; in the isolated-singularity case this forces φ to be an automorphism globally.
- The quasi-finite strategy provides a purely algebraic route to the smooth case, avoiding the analytic methods of earlier proofs.
- Since reducing to normal varieties is already available, these cases cover all algebraic surfaces and all smooth varieties in full generality.
Reading between the lines
- If the Kallström criterion extends, the locally complete intersection case may need no regularity-in-codimension-2 hypothesis; the reflexivity argument only uses (R2) to make Ω_X reflexive, and other normal l.c.i. varieties can fail reflexivity in smaller codimension.
- The smooth-locus automorphism conclusion for Q-factorial varieties suggests a sheaf-theoretic route toward proving the full conjecture for all Q-factorial normal varieties by obtaining φ^{-1}(W)=W from a stronger purity statement.
- A natural computational testbed is toric varieties, where Kähler differentials and codimension conditions are combinatorial; verifying the conjecture for toric threefolds with isolated singularities would exercise Theorem 1.4 on a large, explicit family.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies Miyanishi's conjecture: an endomorphism of an algebraic variety over an algebraically closed field of characteristic zero is an automorphism if it is injective outside a closed subset of codimension at least 2. The authors prove the conjecture when the variety is non-singular, when it is a surface, and when it is locally a complete intersection regular in codimension 2. They also prove auxiliary results showing that, under several hypotheses, the endomorphism restricts to an automorphism of the smooth locus, and that this smooth-locus condition suffices for the full conjecture when singularities are isolated. The proofs are largely algebro-geometric, using Kähler differentials, reflexivity arguments, Zariski's Main Theorem, and external results of Kallström and Kunz.
Significance. If the arguments are correct, the paper makes substantial progress on a long-standing conjecture, giving new cases beyond the previously known affine/complete and non-singular settings. The non-singular case is proved by a clean sheaf-theoretic argument rather than the analytic methods used earlier, and the surface case is elementary. The paper is honest about its dependencies: the main new cases rest on Kallström's results, and the authors have included a reproof of the quasi-finite case, so the self-citation to Das's earlier work is not circular. However, the l.c.i. and Q-factorial cases are only as strong as the verification of the hypotheses of Kallström's corollary, and that verification is incomplete in the manuscript.
major comments (3)
- [Lemma 4.8 and its applications in §§4.2, 5] The proof of Lemma 4.8 verifies only that the source is non-singular and that the target is d.c.i. or satisfies condition (W), then invokes [14, Corollary 4.9] without listing the remaining hypotheses. In the application, the morphism is φ3 = φ|_W, the restriction of an arbitrary endomorphism to the smooth locus, and such a morphism is not proper in general; for example, (x,y) ↦ (x,xy) on A^2 is birational but not proper, and its restriction to the smooth locus has the same failure. If [14, Corollary 4.9] requires properness or finiteness, Lemma 4.8 is not applicable, and Theorems 1.2(iii), 1.3(ii)–(iii), and 4.2(2)–(3) do not follow. The authors should either state and verify every hypothesis of [14, Corollary 4.9] in the stated generality, or prove Lemma 4.8 directly.
- [Proposition 4.3, used in Theorem 4.2(1)] The proof of Proposition 4.3 applies the Zariski–Van der Waerden purity theorem to the restriction φ2 : φ^{-1}(W) → W, but φ2 is not known to be proper or finite. The cited form of purity in [21, Chapter III.9] requires such a hypothesis, and the codimension-1 conclusion for the minimal exceptional locus is otherwise unsupported. This step is load-bearing for Theorem 4.2(1). In the same proof, the claim that φ2 is an isomorphism in the first case of Theorem 4.2 is not immediate from Proposition 4.3 alone; the surjectivity hypothesis of Theorem 4.2(1) must be used explicitly to show that the image of the open immersion is all of W.
- [Proof of Theorem 4.2(4), application of [14, Theorem 4.6]] The fourth case of Theorem 4.2 applies [14, Theorem 4.6] to the restriction φ3 : W → X, which is again a non-proper morphism in general. The manuscript checks the stated conditions (F), the vanishing of ramification on W, and the stalk-level conditions, but it does not verify that [14, Theorem 4.6] is valid for arbitrary birational morphisms rather than proper or finite ones. Since this is an external theorem whose hypotheses are not reproduced, the reader cannot certify the step. Please provide the full statement of [14, Theorem 4.6] and a point-by-point verification of all of its hypotheses.
minor comments (5)
- [Convention 2.3 and Remark 2.2] The notation U, V, Z, S, W is used consistently, but the definition of Z as X \ Image(φ1) could be stated more explicitly as the reduced closed complement, since φ1 is an open immersion.
- [Proof of Lemma 2.4] The argument that the open immersion ι is proper because φ2∘ι is an isomorphism would be clearer if the separatedness of φ2 were mentioned explicitly; this is implicit but not stated.
- [Proof of Lemma 5.1] The assertion that Tor^R_1(S,M) is a torsion R-module is correct because M is finitely generated and S is a local domain flat over R, but this justification is omitted; a parenthetical reference would help.
- [Theorem 5.3] The proof invokes Theorem 1.3 to conclude that φ restricts to an automorphism of W, but does not note that in the l.c.i. case this is exactly Theorem 1.3(iii), whose proof depends on Lemma 4.8 and hence on the verification requested above. This dependency should be made explicit.
- [Notation in §4.2] In the proof of the fourth case of Theorem 4.2, the phrase 'height' is used interchangeably with 'codimension'; since the local rings are normal, this is standard, but a short parenthetical would avoid confusion.
Circularity Check
No significant circularity: the proofs are self-contained modulo external theorems (Ax, Zariski Main Theorem, Kallström purity theorem) and do not assume Miyanishi's conjecture.
full rationale
The paper's derivation chain is not circular. The main cases are proved by reducing the endomorphism to a quasi-finite birational map and then invoking Ax's theorem, an external result, to conclude automorphism. The nonsingular and surface cases are proved directly from Kähler differentials and quasi-finiteness. The l.c.i. and Q-factorial cases rely on Lemma 4.8, which imports Kallström's purity result [14, Corollary 4.9]; this is an external theorem, and the paper verifies its hypotheses as it understands them. The skeptical concern about properness is a correctness risk about whether Kallström's hypotheses are satisfied, not a circularity: the conclusion is not assumed as an input. The self-citations to Das [6] are introductory or auxiliary and are not load-bearing: the quasi-finite lemma is reproved in Lemma 2.7 using Zariski's Main Theorem and Ax's theorem, and the nonsingular case is proved independently in Theorem 3.1. No fitted parameter is relabeled as a prediction, no definition is chosen in terms of the target conclusion, and no central claim reduces to a self-citation chain. Therefore the circularity score is 0.
Assumptions & free parameters
assumptions (7)
- standard math Ax's theorem: an injective endomorphism of an algebraic variety is an automorphism.
- standard math Zariski's Main Theorem: a birational quasi-finite morphism of varieties is an open immersion.
- standard math Jarden's theorem (Prop 2.6): Y and Z = X \ φ(X\Y) have equal dimension.
- standard math Zariski-Van der Waerden purity theorem: the exceptional locus of a birational morphism between normal varieties has codimension 1 if nonempty.
- standard math Results of Kallström (Corollary 4.9, Theorem 4.6, Proposition 2.13, Remark 4.7 of [14])
- standard math Auslander-Buchsbaum formula and Serre's conditions (R_k), (S_k)
- standard math Lipman [19, Prop 8.1]: For an l.c.i. variety satisfying (R_2), the sheaf of Kähler differentials is reflexive.
Cite this review
Pith. "Pith review of On the injective self-maps of algebraic varieties." pith.science (2026). https://pith.science/paper/LOYMBZ6H
@misc{pith2026250418488,
author = {Pith},
title = {Pith review of: On the injective self-maps of algebraic varieties},
year = {2026},
howpublished = {\url{https://pith.science/paper/LOYMBZ6H}},
note = {Machine review of arXiv:2504.18488}
}
abstract
A conjecture of Miyanishi says that an endomorphism of an algebraic variety, defined over an algebraically closed field of characteristic zero, is an automorphism if the endomorphism is injective outside a closed subset of codimension at least $2$. We prove the conjecture in the following cases: (1) The variety is non-singular. (2) The variety is a surface. (3) The variety is locally a complete intersection that is regular in codimension $2$. We also discuss a few instances where an endomorphism of a variety, satisfying the hypothesis of the conjecture of Miyanishi, induces an automorphism of the non-singular locus of the variety. Under additional hypotheses, we prove that the conjecture holds when the variety has only isolated singularities.
Reference graph
Works this paper leans on
-
[1]
Injective endomorphisms of varieties and schemes
Ax, J. Injective endomorphisms of varieties and schemes. Pacific J. Math. 31 (1969), 1–7
work page 1969
-
[2]
Bia/suppress lynicki Birula, A., and Rosenlicht, M.Injective morphisms of real algebraic varieties. Proc. Amer. Math. Soc. 13 (1962), 200–203
work page 1962
-
[3]
Injective endomorphisms of algebraic varieties
Borel, A. Injective endomorphisms of algebraic varieties. Arch. Math. (Basel) 20 (1969), 531–537
work page 1969
-
[4]
Bruns, W., and Herzog, J. Cohen-Macaulay rings , vol. 39 of Cambridge Studies in Advanced Mathematics. Cambridge University Press, Cambridge, 1993
work page 1993
-
[5]
Injective endomorphisms of algebraic and analytic sets
Cynk, S., and Rusek, K. Injective endomorphisms of algebraic and analytic sets. Ann. Polon. Math. 56 , 1 (1991), 29–35
work page 1991
-
[6]
On endomorphisms of algebraic varieties
Das, N. On endomorphisms of algebraic varieties. Int. Math. Res. Not. IMRN , 22 (2022), 17534– 17545
work page 2022
-
[7]
Endomorphisms of symbolic algebraic varieties
Gromov, M. Endomorphisms of symbolic algebraic varieties. J. Eur. Math. Soc. (JEMS) 1 , 2 (1999), 109–197
work page 1999
-
[8]
´El´ ements de g´ eom´ etrie alg´ ebrique
Grothendieck, A. ´El´ ements de g´ eom´ etrie alg´ ebrique. IV.´Etude locale des sch´ emas et des mor- phismes de sch´ emas IV.Inst. Hautes ´Etudes Sci. Publ. Math. , 32 (1967), 361
work page 1967
Show all 24 references
-
[9]
Algebraic geometry
Hartshorne, R. Algebraic geometry . Graduate Texts in Mathematics, No. 52. Springer-Verlag, New York-Heidelberg, 1977
1977
-
[10]
Stable reflexive sheaves
Hartshorne, R. Stable reflexive sheaves. Math. Ann. 254 , 2 (1980), 121–176
1980
-
[11]
On logarithmic Kodaira dimension of algebraic varieties
Iitaka, S. On logarithmic Kodaira dimension of algebraic varieties. In Complex analysis and alge- braic geometry. Iwanami Shoten, Tokyo, 1977, pp. 175–189
1977
-
[12]
An injective rational map of an abstract algebraic variety into itself
Jarden, M. An injective rational map of an abstract algebraic variety into itself . J. Reine Angew. Math. 265 (1974), 23–30
1974
-
[13]
On a theorem of Ax
Kaliman, S. On a theorem of Ax. Proc. Amer. Math. Soc. 133 , 4 (2005), 975–977
2005
-
[14]
Purity of branch and critical locus
K¨allstr¨om, R. Purity of branch and critical locus. J. Algebra 379 (2013), 156–178
2013
-
[15]
K¨ ahler differentials
Kunz, E. K¨ ahler differentials. Advanced Lectures in Mathematics. Friedr. Vieweg & Sohn, Braun - schweig, 1986
1986
-
[16]
Injective endomorphisms of real algebraic sets are surjective
Kurdyka, K. Injective endomorphisms of real algebraic sets are surjective. Math. Ann. 313 , 1 (1999), 69–82
1999
-
[17]
Surjectivity of certain injective semialgebraic transformations of Rn
Kurdyka, K., and Rusek, K. Surjectivity of certain injective semialgebraic transformations of Rn. Math. Z. 200 , 1 (1988), 141–148. 14 I. BISW AS AND N. DAS
1988
-
[18]
Birational self-maps and piecewise algebraic geometry
Lamy, S., and Sebag, J. Birational self-maps and piecewise algebraic geometry. J. Math. Sci. Univ. Tokyo 19 , 3 (2012), 325–357
2012
-
[19]
Free derivation modules on algebraic varieties
Lipman, J. Free derivation modules on algebraic varieties. Amer. J. Math. 87 (1965), 874–898
1965
-
[20]
Open problems
Miyanishi, M. Open problems. In Affine algebraic geometry , vol. 369 of Contemp. Math. Amer. Math. Soc., Providence, RI, 2005, pp. 10–12
2005
-
[21]
The red book of varieties and schemes , expanded ed., vol
Mumford, D. The red book of varieties and schemes , expanded ed., vol. 1358 of Lecture Notes in Mathematics. Springer-Verlag, Berlin, 1999. Includes the Michigan lectures (19 74) on curves and their Jacobians, With contributions by Enrico Arbarello
1999
-
[22]
Nowak, K. J. Injective endomorphisms of algebraic varieties. Math. Ann. 299 , 4 (1994), 769–778
1994
-
[23]
Topology of injective endomorphisms of real algebraic sets
Parusi´nski, A. Topology of injective endomorphisms of real algebraic sets. Math. Ann. 328 , 1-2 (2004), 353–372
2004
-
[24]
The stacks project
Stacks project authors, T. The stacks project. https://stacks.math.columbia.edu, 2023. Mathematics Department, Shiv Nadar University, NH91, Tehsil Dadri, Greater Noida, Uttar Pradesh 201314, India Email address : indranil.biswas@snu.edu.in, indranil29@gmail.com Stat-Math Unit,...
2023
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.