{"id":"c47fa7ca-8c7b-4043-9d21-f9c9b340ed7e","arxiv_id":"1908.02117","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Open embeddings of Stein spaces and of C∞-manifolds are exactly the maps whose induced homomorphism of function algebras is a 1-pseudoflat epimorphism, with equivalent homological conditions in the smooth case.","lead":"In mathematics, this paper proves that 'open embedding' of two kinds of spaces is exactly the same as a 'pseudoflat epimorphism' between their algebras of functions. It links geometry to algebraic conditions that can be checked with homological algebra.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.3's (vii)⇒(i) rests on an unproved extension of Ogneva's projectivity theorem to arbitrary open subsets; if that extension fails, the projective/flat/strong homological equivalences are unsupported.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing point: Theorem 5.3 (vii)⇒(i) depends on an unverified extension of Ogneva's theorem, admitted in the paper's own footnote. I agree that this is the weakest link in the central argument. The concern is structural, not a matter of disagreement with consensus: even if the extension is in fact true, the paper does not supply the proof, and the natural arguments one might use involve subtleties about products versus direct sums of projective Fréchet modules. The manuscript's own footnote is therefore an in-scope limitation that should be flagged. Since the reader already assigned a CONDITIONAL verdict for this reason, my read does not move the verdict; I recommend keeping it as is. The auxiliary issue around Theorem 3.24 and condition (∗) is real but does not feed into Theorems 4.2 or 5.3, so it is not the most load-bearing concern for the central claim.","tokens_in":22196,"tokens_out":29174,"duration_ms":310179,"concrete_test":"Obtain [37] and verify the precise statement and proof of the projectivity of C∞(Y) over C∞(X) for Y contained in a coordinate neighborhood. Then test the generalization in a case where Y is not contained in a single chart, e.g., X = R×S^1 and Y an open subset that is not contained in any chart and requires an infinite locally finite atlas. Attempt to prove projectivity of C∞(Y) over C∞(X) using a partition of unity; check whether the construction yields a direct summand of a finite direct sum of free modules, or only of an infinite product of projectives. If only an infinite product is obtained, find an admissible epimorphism onto C∞(Y) that does not split; such an epimorphism would disprove the claimed generalization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing point is in Theorem 5.3, proof of (vii)⇒(i). To show that an open embedding f:Y→X induces a projective epimorphism, the paper invokes Ogneva's theorem [37] that C∞(Y) is projective over C∞(X). The attached footnote admits that [37] proves this only when Y is contained in a coordinate neighborhood of X, and states without proof that the argument 'readily carries over' to the general case. This matters because (i) is the entry point of the implication chain (i)⇒(ii)⇒(iv)⇒(v); if the generalization is false, the equivalence between open embeddings and projective, flat, strong homological, and weak homological epimorphisms in Theorem 5.3 is not established. The extension is not formally automatic: a natural locally finite partition-of-unity argument would realize C∞(Y) as a direct summand of a product of the modules C∞(U_i∩Y) (with U_i coordinate neighborhoods), but an infinite product of projective Fréchet modules need not be projective in Helemskii's category, so projectivity does not follow without an additional argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies continuous homomorphisms of Fréchet algebras induced by maps of Stein spaces and of smooth manifolds. It introduces n-pseudoflat epimorphisms, defined by vanishing of the first n Tors and by the canonical isomorphism at Tor_0, and proves that for Stein spaces a holomorphic map is an open embedding exactly when the induced map on holomorphic function algebras is a 1-pseudoflat epimorphism (Theorem 4.2). For smooth manifolds it claims a stronger equivalence: open embeddings are characterized by projective, flat, strong homological, weak homological, and 1-pseudoflat epimorphisms (Theorem 5.3). The paper also corrects a gap in an earlier version of Theorem 3.24 by adding a condition (*) to the differential-form characterization of 1-pseudoflat epimorphisms.","tokens_in":22398,"tokens_out":3564,"duration_ms":36482,"significance":"The main results are natural and significant: they provide analytic and smooth analogues of the algebraic characterization of open embeddings as flat epimorphisms of finite presentation, and they do so in a setting where ordinary flatness is known to fail. The paper is carefully written and gives substantial historical context. A particular strength is that the authors explicitly acknowledge and repair the previous gap in Theorem 3.24. However, the smooth case rests on an external projectivity theorem of Ogneva whose extension to arbitrary open subsets is asserted only in a footnote. Since this extension is used as the entry point of the implication chain (vii)⇒(i) in Theorem 5.3, the full equivalence is not yet established as written. The gap appears fixable, but it must be addressed before the central claim can be accepted.","major_comments":[{"comment":"The proof of (vii)⇒(i) invokes [37, Theorem 2] to conclude that C∞(Y) is projective over C∞(X) for an arbitrary open subset Y of X, with the footnote stating that the proof in [37], given for Y contained in a coordinate neighborhood, 'readily carries over' to the general case. This is a load-bearing step: it is the only argument for projectivity, which then feeds into the equivalences with flat, strong homological, and weak homological epimorphisms. The extension is not automatic. A natural reduction using a locally finite partition of unity would express C∞(Y) as a direct summand of a product of modules C∞(Ui∩Y), but an infinite product of projective Fréchet modules need not be projective in Helemskii's category of Fréchet modules. The authors should either supply a complete proof of the extension or give a precise reference that covers the general case; the current footnote is insufficient.","section":"Theorem 5.3, proof of (vii)⇒(i), footnote 1"}],"minor_comments":[{"comment":"The proof of Lemma 5.1 is omitted as standard. Although this is a routine hom-tensor adjunction, the lemma is used in the proof of Lemma 5.2, which is itself used in the proof of (vi)⇒(vii) of Theorem 5.3. The authors should either include the short proof or cite a published statement that covers the Fréchet module setting.","section":"Lemma 5.1"},{"comment":"The dedication contains a typo: 'occas ion' should be 'occasion'.","section":"Dedication and text"},{"comment":"The note explaining that condition (*) was missing in the first version and in the published journal version is helpful and honest, but the statement that the authors do not know whether (*) is essential leaves a small unresolved question; it would be good to add a remark on whether (*) can be verified in the main geometric examples without additional work.","section":"Theorem 3.24, footnote about condition (*)"},{"comment":"The comparison with flatness results in [1] and [3] is interesting, but the sentence 'the restriction map O(C)→O(D) is a 1-pseudoflat epimorphism by Theorem 4.2' might give the impression of circularity, since Theorem 4.2 is the paper's own result; the immediate reference to [52, Prop. 3.1] already placed in parentheses resolves this, so the wording could be tightened.","section":"Remark 4.3"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the unproved extension of Ogneva's theorem in Theorem 5.3. This is a genuine gap in the manuscript as written, but it is localized and likely repairable. The authors' explicit disclosure of the earlier gap in Theorem 3.24 is commendable and should not be held against them. I recommend major revision rather than rejection: the central ideas are sound and the missing argument appears to be within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Solid homological characterization of open embeddings; the smooth-case equivalence is conditional on an unproved extension of Ogneva's projectivity theorem. The paper gives a clean homological characterization of open embeddings for Stein spaces and C-infinity manifolds. The main theorems are Theorem 4.2 and Theorem 5.3, and I think they are probably right, but one step in the smooth case is genuinely under-supported: the proof of (vii)=> (i) in Theorem 5.3 depends on the assertion that Ogneva's projectivity theorem for C-infinity(Y) over C-infinity(X), proved only for Y in a coordinate neighborhood, 'readily carries over' to arbitrary open subsets. No argument is given. A natural partition-of-unity approach would write C-infinity(Y) as a direct summand of an infinite product of the C-infinity(U_i cap Y), but arbitrary products of projective Frechet modules need not be projective, so this is not automatic. If the extension fails, the equivalences with projective, flat, and strong homological epimorphisms in Theorem 5.3 are not established. The remaining implications, (ii) through (vii), look solid and give a robust equivalence between open embeddings and 1-pseudoflat epimorphisms, which is the core of the paper. What is genuinely new: the introduction of n-pseudoflat epimorphisms in the Frechet setting, the characterization via noncommutative differential forms, and the smooth-manifold theorem. The Stein result is explicitly a partial generalization of Bambozzi-Ben-Bassat-Kremnizer, which is fine; the Frechet-topological version is not in the literature. The paper is also honest about its own history: the gap in Theorem 3.24 is flagged, the corrected condition (*) is introduced with an explicit statement that its necessity is unknown, and a corrigendum is promised. That is careful scholarship. The other soft spots are minor. Theorem 3.24's condition (*) is a bit awkward but it does not feed into the main theorems. Lemma 5.1's proof is omitted as standard, which is acceptable. The citation pattern is thorough and not self-serving. Who is this for: specialists in topological homology and analytic geometry. It deserves a serious referee, but the referee should press for a proof of the Ogneva extension, or a reformulation that avoids it. If the extension turns out false, the paper's main characterization of 1-pseudoflat epimorphisms still stands, but the stronger homological conditions need to be separated. Send it to review.","headline":"Solid homological characterization of open embeddings; the smooth-case equivalence is conditional on an unproved extension of Ogneva's projectivity theorem.","tokens_in":706,"tokens_out":1666,"would_cite":true,"duration_ms":62194,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46M18","46H25","46E25","32A38","16E30"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that open embeddings of Stein spaces and of C∞-manifolds are exactly the 1-pseudoflat epimorphisms induced on their function algebras.","keywords":["open embeddings","Stein spaces","C∞-manifolds","pseudoflat epimorphisms","homological epimorphisms","Fréchet algebras","Tor functor","function algebras"],"falsifier":"Take a smooth compact manifold such as the two-dimensional torus, remove two disjoint closed disks so that the remaining open set is not contained in any coordinate chart, and check whether the smooth functions on it still form a projective Fréchet module over the smooth functions on the torus; if not, the theorem's implication from open embedding to projective epimorphism fails.","tokens_in":21959,"feed_emoji":"📐","tokens_out":13620,"duration_ms":176539,"temperature":0.7,"pith_summary":"This paper sets out to find a homological condition on the algebra of holomorphic or smooth functions that exactly detects when a map between spaces is an open embedding. For Stein spaces it proves that a holomorphic map $f:Y\\to X$ is an open embedding if and only if the induced homomorphism $f^\\bullet:\\mathcal{O}(X)\\to\\mathcal{O}(Y)$ is a 1-pseudoflat epimorphism, and for $C^\\infty$-manifolds the corresponding statement holds with the additional equivalence to flat and projective epimorphisms. This matters because in the algebraic setting open embeddings of affine schemes are known to be the flat epimorphisms of finite presentation, while the analytic version of flatness fails: the algebra of holomorphic functions on an open subset is usually not flat over the whole algebra. The paper's substitute notion replaces full flatness by the vanishing of only the first Tor and a canonical zeroth-Tor identification, which is strong enough to force local biholomorphy or local diffeomorphy at every point.","feed_headline":"Open embeddings have one algebraic test","feed_subtitle":"For Stein holomorphic maps and smooth maps, open embeddings are exactly the 1-pseudoflat epimorphisms of function algebras.","key_machinery":"The key object is the $n$-pseudoflat epimorphism: a continuous homomorphism $\\phi:A\\to B$ of Fréchet algebras such that $\\operatorname{Tor}^A_i(B,B)=0$ for $1\\le i\\le n$ and $\\operatorname{Tor}^A_0(B,B)\\cong B$ canonically. The paper's main technical tool is the characterization (Theorem 3.24) of 1-pseudoflat epimorphisms in terms of noncommutative differential forms and derivations: $\\phi$ is a 1-pseudoflat epimorphism exactly when the induced map $\\operatorname{Der}(B,X)\\to\\operatorname{Der}(A,X)$ is bijective for every Fréchet $B$-bimodule $X$ and an extra Hausdorffness condition $(*)$ holds. This differential-form criterion is what connects the abstract Tor conditions to geometry: applied to point modules $\\mathbb{C}_q$, it forces the tangent map of $f$ to be bijective, while epimorphicity plus the character-space bijection for Stein spaces and manifolds forces $f$ to be injective. Together these yield local biholomorphy or local diffeomorphy, and hence openness of the embedding.","core_discovery":"The central discovery is an exact dictionary between geometry and homological algebra. A morphism $f:Y\\to X$ of Stein spaces is an open embedding exactly when $f^\\bullet:\\mathcal{O}(X)\\to\\mathcal{O}(Y)$ is a 1-pseudoflat epimorphism, equivalently a weak homological epimorphism, equivalently an epimorphism satisfying the pointwise transversality condition $\\mathcal{O}(Y)\\perp^1_{\\mathcal{O}(X)}\\mathbb{C}_q$ for every $q\\in Y$. For smooth manifolds the same conclusion holds, and the 1-pseudoflat condition is also equivalent to $f^\\bullet$ being a projective epimorphism, a flat epimorphism, a strong homological epimorphism, or a weak homological epimorphism. In the Stein case flatness is genuinely too strong to be the right test, so the 1-pseudoflat epimorphism is the correct analytic replacement.","pith_inferences":["If the Stein restriction maps beyond polydomains are also strong homological epimorphisms, the Stein and smooth theorems would match closely; the open unit ball in $\\mathbb{C}^2$ is the concrete test case the paper leaves open.","The derivation bijection in Theorem 3.24 suggests interpreting 1-pseudoflatness as a first-order infinitesimal test: applying it at point modules gives the tangent-space isomorphisms that force local diffeomorphy, so replacing point modules by jet modules could yield homological tests for submersions or immersions.","Since a quotient map onto a Whitney subspace is 1-pseudoflat without being an open embedding, a $C^\\infty$-differentiable-space analogue of Theorem 5.3 will need an additional injectivity or separation condition beyond Tor vanishing."],"forward_implications":["A holomorphic map between Stein spaces is an open embedding exactly when its induced algebra homomorphism is a 1-pseudoflat epimorphism, so the geometry of the map can be read from the vanishing of the first Tor space and a canonical zeroth-Tor identification.","For smooth maps between $C^\\infty$-manifolds the same 1-pseudoflat test also implies flatness, projectivity, and strong homological-epimorphism properties of the induced homomorphism, giving several independent algebraic certificates for openness.","The classical affine-scheme theorem that flat epimorphisms of finite presentation are open embeddings now has analytic and smooth analogues in which finite presentation is not needed and flatness is weakened to 1-pseudoflatness.","Every restriction map $\\mathcal{O}(X)\\to\\mathcal{O}(U)$ to a Stein open subset is a weak homological epimorphism, so the localization framework covers all Stein open subsets, not just polydomains in $\\mathbb{C}^n$."],"supporting_citations":[{"why":"Supplies the affine-scheme theorem that flat epimorphisms of finite presentation are open embeddings, the algebraic prototype being adapted here.","marker":"[26]"},{"why":"Gives the bornological-algebra characterization of open embeddings of Stein spaces that this paper partially generalizes to topological Fréchet algebras.","marker":"[2]"},{"why":"Introduces the localization and absolute-localization conditions and proves the restriction map to a Stein open subset of $\\mathbb{C}^n$ satisfies them, also providing the polydomain projectivity used later.","marker":"[52]"},{"why":"Extends the weak-homological-epimorphism result to Stein manifolds and shows the unit-disc restriction is not flat, motivating the pseudoflat substitute.","marker":"[38]"},{"why":"Provides the transversality relation $\\mathcal{O}(Y)\\perp_{\\mathcal{O}(X)}F(Z)$ and the Tor machinery used in the open-embedding direction of the Stein theorem.","marker":"[16]"},{"why":"Supplies the projectivity of $C^\\infty(Y)$ over $C^\\infty(X)$ for open subsets, which closes the smooth implication (vii)$\\Rightarrow$(i); the paper notes the stated proof covers only coordinate neighborhoods.","marker":"[37]"},{"why":"Gives the bijection between points of a Stein space and characters of its function algebra, used to prove $f$ is injective.","marker":"[18]"},{"why":"Gives the analogous character-space bijection for smooth manifolds, used to prove injectivity of $f$ in the $C^\\infty$ case.","marker":"[36]"}],"fun_headline_variants":["Pseudoflat epimorphisms detect open embeddings","One algebraic test for open embeddings","When flatness is too strong, pseudoflat works","Open embeddings are 1-pseudoflat epimorphisms","A new flatness-type condition for open embeddings"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The smooth case depends on a cited result asserting that for every open subset of a manifold, the smooth function algebra on that subset behaves homologically like a direct factor of the smooth function algebra on the whole manifold; the cited proof covers only open subsets lying inside a coordinate chart, and the paper asserts the general case without giving the argument.","fun_headline_variants_meta":{"raw":{"variants":["Pseudoflat epimorphisms detect open embeddings","One algebraic test for open embeddings","When flatness is too strong, pseudoflat works","Open embeddings are 1-pseudoflat epimorphisms","A new flatness-type condition for open embeddings"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000163,"raw_usage":{"total_tokens":1140,"prompt_tokens":739,"completion_tokens":401,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":355,"completion_tokens_details":{"reasoning_tokens":328}},"tokens_in":355,"tokens_out":401,"duration_ms":4663,"temperature":1.0,"reasoning_tokens":328,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:53:11.049271+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a smooth compact manifold such as the two-dimensional torus, remove two disjoint closed disks so that the remaining open set is not contained in any coordinate chart, and check whether the smooth functions on it still form a projective Fréchet module over the smooth functions on the torus; if not, the theorem's implication from open embedding to projective epimorphism fails.","supporting_citations":[{"cited_title":"´El´ ements de g´ eom´ etrie alg´ ebrique","cited_arxiv_id":null,"evidence_quote":"Supplies the affine-scheme theorem that flat epimorphisms of finite presentation are open embeddings, the algebraic prototype being adapted here."},{"cited_title":"Stein domains in Banach algebraic geometry","cited_arxiv_id":null,"evidence_quote":"Gives the bornological-algebra characterization of open embeddings of Stein spaces that this paper partially generalizes to topological Fréchet algebras."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the localization and absolute-localization conditions and proves the restriction map to a Stein open subset of $\\mathbb{C}^n$ satisfies them, also providing the polydomain projectivity used later."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the weak-homological-epimorphism result to Stein manifolds and shows the unit-disc restriction is not flat, motivating the pseudoflat substitute."},{"cited_title":"and Putinar, M","cited_arxiv_id":null,"evidence_quote":"Provides the transversality relation $\\mathcal{O}(Y)\\perp_{\\mathcal{O}(X)}F(Z)$ and the Tor machinery used in the open-embedding direction of the Stein theorem."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the projectivity of $C^\\infty(Y)$ over $C^\\infty(X)$ for open subsets, which closes the smooth implication (vii)$\\Rightarrow$(i); the paper notes the stated proof covers only coordinate neighborhoods."},{"cited_title":"Zur Theorie der Steinschen Algebren und Moduln , Math","cited_arxiv_id":null,"evidence_quote":"Gives the bijection between points of a Stein space and characters of its function algebra, used to prove $f$ is injective."},{"cited_title":"Smooth manifolds and observables","cited_arxiv_id":null,"evidence_quote":"Gives the analogous character-space bijection for smooth manifolds, used to prove injectivity of $f$ in the $C^\\infty$ case."}],"review_version":1}