{"id":"9aead101-8855-4c2f-96ef-6a766be951be","arxiv_id":"1908.07257","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The prime and maximal spectra of rings of C^r semialgebraic functions are homeomorphic to those of the ring of continuous semialgebraic functions, which is their real closure.","lead":"This paper studies smooth functions on shapes defined by polynomial equations and shows that all smoothness levels share the same algebraic skeleton as the continuous functions on those shapes. It matters because it completes the structural theory of these function rings and clarifies their relation to Nash functions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Corollary 5.9 overstates its support: integral closedness in condition (iii.a) is not a consequence of Theorem 1.4, so the claim that S^r(M) fails only property (iv) is underevidenced.","rationale":"The reader's stated weakest assumption, Fact 2.4, is an external theorem; it is the natural bridge for the Whitney-extension part of the argument, and I found no reason to doubt it. The paper's central spectral homeomorphism is detailed and internally consistent as far as I checked. The genuinely load-bearing weakness lies in a consequence of the main result that the authors advertise prominently: Corollary 5.9 and Remark 5.2 claim that S^r(M) satisfies all real-closed-ring axioms except the sum-of-radical-ideals condition. The proof of condition (iii.a) is a single unsupported sentence. Since a domain whose fraction field is real closed need not be integrally closed in it, Theorem 1.4 cannot by itself justify that sentence. The reader's rationale already noted this gap, though the formal 'weakest assumption' field points to Fact 2.4 instead. I therefore partially agree with the reader. The verdict should remain conditional: Theorem 1.2 and the real-closure statement Proposition 5.7 are not undermined, but the paper's claim about the exact status of S^r(M) relative to real closed rings needs a repaired proof or a revised statement.","tokens_in":50929,"tokens_out":38246,"duration_ms":398631,"concrete_test":"Independently prove or disprove the missing assertion: for every semialgebraic M and every q in Spec S^r(M), the domain S^r(M)/q is integrally closed in qf(S^r(M)/q). A decisive first test is to compute this for M = R, r = 1 and for a prime q coming from a non-archimedean ordering of R(t), for example the kernel of the natural map from S^1(R) to an ultrapower or to a real closure of R(t) that makes x positive infinitesimal. Check whether the class of x^{1/2}, represented as x^{3/2}/x, lies in the fraction field and satisfies X^2 - x = 0 while not lying in S^1(R)/q. If such a q exists, Corollary 5.9(iii.a) is false; if it provably does not exist, the paper still needs to state the missing convexity or Sper-constructibility argument that proves integral closedness.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central Theorem 1.2 appears well supported: the proof through the Lojasiewicz Nullstellensatz for locally compact sets, the weak Nullstellensatz plus direct limits for general sets, and localization/Nash approximation for the bounded case is coherent, and I do not share the reader's doubt about Fact 2.4, which is the known o-minimal Whitney extension theorem from [Th]/[KP2]. The real gap is in Corollary 5.9 and Remark 5.2. The text says 'Condition (iii.a) follows from Theorem 1.4.' Theorem 1.4 proves only that the fraction field of S^r(M)/q is a real closed field. That alone does not imply that S^r(M)/q is integrally closed in that fraction field. For example, in a non-archimedean real closed field K with convex-hull valuation ring O = R + M and maximal ideal M, the subring R + M^2 has fraction field K but is not integrally closed: any x in M \\ (R + M^2) satisfies the monic equation X^2 - x^2 = 0 with x^2 in R + M^2. So integral closedness is an additional property, not a formal consequence of real-closedness of the fraction field. The proof of Corollary 5.9 supplies no supplementary argument for it. Thus the advertised conclusion that S^r(M) satisfies conditions (i)-(iii) of Definition 5.1 and misses only property (iv) is not established. This does not invalidate Theorem 1.2, but it weakens the real-closed-ring interpretation of the spectra homeomorphism.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops an intrinsic theory of differentiable semialgebraic functions of class C^r on arbitrary semialgebraic sets, following the jet-based definition of Aschenbrenner and Thamrongthanyalak. Its central result, Theorem 1.2, asserts that for every semialgebraic set M and every r≥0, the restriction map φ: Spec^{0⋄}(M)→Spec^{r⋄}(M), p↦p∩S^{r⋄}(M), is a homeomorphism whose inverse is q↦√(qS^0(M)), and that this homeomorphism restricts to the maximal spectra. The proof is organized by cases: locally compact sets via a Lojasiewicz Nullstellensatz for S^r-functions (Theorem 3.2), general sets via a weak Nullstellensatz and direct limits, and bounded functions via Nash approximation. The paper further proves Theorem 1.4, that the residue field of every prime ideal of S^{r⋄}(M) is real closed, and proposes in Section 5 that S^{r⋄}(M) satisfies all conditions of Schwartz's definition of a real closed ring except the sum of radical ideals being radical. It also compares S^∞(M) with the ring of Nash functions N(M), proving local Nashness of S^∞ functions and a criterion for equality on coherent Nash sets.","tokens_in":51033,"tokens_out":6624,"duration_ms":72567,"significance":"If the main results are correct, the paper gives a substantial and surprising rigidity theorem: the Zariski and maximal spectra of the rings of C^r semialgebraic functions are independent of r and coincide with those of the real closed ring S^0(M). The explicit inverse √(qS^0(M)) is valuable, and the consequences—Gelfand property, Krull dimension equal to dim(M), homeomorphic maximal spectra—are natural and strong. The proof strategy is coherent: it rests on the known o-minimal Whitney extension theorem of Thamrongthanyalak (Fact 2.4), the authors' earlier work on rings of semialgebraic functions, Schwartz's real closed ring theory, and Nash approximation. I do not share the reader's doubt about Fact 2.4; it is a standard external result, not a source of circularity. The direct-limit presentation and the brimming-completion argument in Section 4 are careful and detailed. However, one advertised consequence in Section 5, namely that S^{r⋄}(M) satisfies condition (iii.a) of Definition 5.1, is not established by the proof given.","major_comments":[{"comment":"The proof of Corollary 5.9 contains the sentence 'Condition (iii.a) follows from Theorem 1.4.' This is not sufficient. Theorem 1.4 states that the inclusion S^{r⋄}(M)/q → S^{0⋄}(M)/p induces an isomorphism of fraction fields, so that κ(q) is real closed. But condition (iii.a) of Definition 5.1 also requires that S^{r⋄}(M)/q be integrally closed in that real closed fraction field. Equality of fraction fields does not imply integral closedness. For example, in a non-archimedean real closed field K with convex-hull valuation ring O=R+M and maximal ideal M, the subring R+M^2 has fraction field K but is not integrally closed: any x∈M∖(R+M^2) satisfies the monic equation X^2−x^2=0 with x^2∈R+M^2. The proof of Corollary 5.9 supplies no supplementary argument that this pathology cannot occur for S^{r⋄}(M)/q. Consequently, the advertised conclusion that S^{r⋄}(M) satisfies conditions (i)–(iii) of Definition 5.1 and fails only property (iv) is not established. This does not invalidate Theorem 1.2, but it weakens the real-closed-ring interpretation of the spectra homeomorphism. Please either prove integral closedness of S^{r⋄}(M)/q in κ(q) or restate the claim to assert exactly what the proof supports.","section":"§5.A, Corollary 5.9"}],"minor_comments":[{"comment":"The text contains many stray '/suppress' tokens, for example in the abstract, in Theorem 1.3, in the headings of Section 3.A, and in the bibliography entries for Lojasiewicz, Pawłucki, and the project title. These appear to be text-extraction artifacts and should be removed before publication.","section":"Throughout"},{"comment":"The Introduction's statement of the Lojasiewicz Nullstellensatz uses Z(f)⊂Z(g), while the later Theorem 3.2 uses Z(f1)⊂Z(f2) with the same meaning; this is harmless but the notation should be unified for readability.","section":"§1, Theorem 1.3"},{"comment":"In the proof of Lemma 5.11, the transition from formal power series to Nash germs via Artin approximation is compressed; a short sentence explaining why the system to which Artin approximation is applied is algebraic over R[x] would help the reader.","section":"§5.B, Lemma 5.11"},{"comment":"The definition of S^r-functions via jets relies on a choice of semialgebraic jet; Remark 2.5 correctly notes non-uniqueness, but it would be useful to state explicitly that S^r(M) is defined as the set of functions admitting at least one such jet.","section":"§2.A, Definition 2.2"}],"recommendation":"major_revision","confidential_remarks":"The central spectral theorem and its proof appear sound and the paper is well within the scope of the journal. The main issue is local but load-bearing for the advertised real-closed-ring interpretation: the proof of Corollary 5.9 does not justify condition (iii.a). If the authors can supply the missing integral-closedness argument, or appropriately weaken the conclusion, I would support acceptance. Please also ensure that the visible '/suppress' artifacts are cleaned from the final version."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe genuinely new results here are the spectral homeomorphisms: for every r ≥ 1, the Zariski and maximal spectra of the rings S^r(M) and S^{r*}(M) of differentiable semialgebraic functions are homeomorphic to those of the continuous ring S^0(M), the residue fields of prime quotients are real closed, and S^0(M) is the real closure of S^r(M). That is a substantial body of theorems. The proof architecture is coherent: the locally compact case via a Lojasiewicz Nullstellensatz, the general case via direct limits, and the bounded case via Nash approximation. I checked the main line and found no errors. I also do not share the reader's doubt about Fact 2.4; the o-minimal Whitney extension theorem from [Th] is a known external result, and using it as the bridge is legitimate.\n\nThe real soft spot is exactly where the stress-test note lands. Corollary 5.9 claims that S^r(M) satisfies conditions (i)–(iii) of Schwartz's definition of a real closed ring and fails only condition (iv). But condition (iii.a) requires that A/q be integrally closed in its fraction field. The proof says only \"Condition (iii.a) follows from Theorem 1.4.\" Theorem 1.4 proves the fraction field is real closed; it does not prove integral closedness. The stress-test's counterexample — R + M^2 inside a non-archimedean real closed field — shows the inference is invalid in general. Perhaps integral closedness is true for these quotient rings for other reasons, but the paper does not provide the argument. This does not damage Theorem 1.2 or the spectral results, but it does undercut the advertised \"close to real closed\" characterization.\n\nMinor issues: the \"problematic point\" and \"brimming completion\" definitions are workable but add technical load. The Nash comparison in Section 5 is interesting but feels secondary.\n\nWho gets value: people working in real algebraic geometry, semialgebraic function rings, and real closed rings. The paper deserves a serious referee. I would send it out with a request that the referee specifically check the real-closed-ring corollary and that the authors either prove integral closedness or soften the claim.\n\nBest,\n[Your name]","headline":"Main spectral homeomorphism theorems are sound and novel; the paper overstates the real-closed-ring corollary by one unproven integral-closedness claim.","tokens_in":51897,"tokens_out":2130,"would_cite":true,"duration_ms":23494,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14P10","46E25","12D15","13E99"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that for every r≥1, the Zariski and maximal spectra of the ring of C^r differentiable semialgebraic functions on a semialgebraic set are homeomorphic to those of the ring of continuous semialgebraic functions, so all…","keywords":["differentiable semialgebraic functions","Zariski spectrum","maximal spectrum","real closed field","real closed ring","real closure","Lojasiewicz Nullstellensatz","Nash functions"],"falsifier":"Construct a semialgebraic set $M$ and a prime ideal $q$ of ${\\mathcal S}^r(M)$ for which $\\sqrt{q{\\mathcal S}^0(M)}\\cap{\\mathcal S}^r(M)$ strictly contains $q$; the paper's inverse formula says these two sets must be equal, so such a pair would falsify Theorem 1.2 and the real-closure statement.","tokens_in":50447,"feed_emoji":"📐","tokens_out":9677,"duration_ms":88604,"temperature":0.7,"pith_summary":"The paper proves that the ring ${\\mathcal S}^r(M)$ of ${\\mathcal C}^r$ differentiable semialgebraic functions on any semialgebraic set $M\\subset\\mathbb{R}^m$ has the same Zariski and maximal spectra as the ring ${\\mathcal S}^0(M)$ of continuous semialgebraic functions on $M$ that extend to an open neighborhood of $M$ in its closure. The same holds for the bounded subrings ${\\mathcal S}^{r*}(M)$ and ${\\mathcal S}^{0*}(M)$. Because the homeomorphism is explicit, every spectral property of the two rings is identical: ${\\mathcal S}^r(M)$ is a Gelfand ring, its Krull dimension equals $\\dim(M)$, and the fraction field of each quotient by a prime ideal is real closed. The paper also shows that ${\\mathcal S}^0(M)$ is the real closure of ${\\mathcal S}^r(M)$, even though ${\\mathcal S}^r(M)$ itself is not real closed because sums of radical ideals need not be radical.","feed_headline":"All C^r semialgebraic function rings share one Zariski spectrum","feed_subtitle":"For every r≥1 the differentiable and continuous semialgebraic rings have homeomorphic spectra, so their algebraic invariants coincide.","key_machinery":"The load-bearing identity is the inverse pair $\\phi(p)=p\\cap{\\mathcal S}^r(M)$ and $\\psi(q)=\\sqrt{q{\\mathcal S}^0(M)}$, which makes the two spectra into the same topological space. The identity is established with three tools: first, the semialgebraic Whitney extension theorem, which says that semialgebraic jets on closed sets extend to ${\\mathcal C}^r$ semialgebraic functions on $\\mathbb{R}^m$ and so supplies the extension results; second, Lojasiewicz's Nullstellensatz for ${\\mathcal S}^r(M)$ on locally compact $M$, which turns zero-set containment into ideal containment and makes radical ideals of ${\\mathcal S}^r(M)$ into $z$-ideals, that is, ideals closed under the rule 'if $Z(f)\\subset Z(g)$ and $f$ is in the ideal, then $g$ is in the ideal'; and third, for the bounded rings, a sup-norm approximation lemma that identifies $\\sqrt{q{\\mathcal S}^{0*}(M)}$ with a prime ideal of ${\\mathcal S}^{0*}(M)$ contracting back to $q$, plus a localization argument reducing to the unbounded case.","core_discovery":"On the paper's own terms, the central discovery is Theorem 1.2: for every semialgebraic set $M\\subset\\mathbb{R}^m$ and every $r\\ge0$, the map $\\phi:\\operatorname{Spec}^{0*}(M)\\to\\operatorname{Spec}^{r*}(M)$, $p\\mapsto p\\cap{\\mathcal S}^{r*}(M)$ is a homeomorphism of Zariski spectra, with inverse $\\psi(q)=\\sqrt{q{\\mathcal S}^{0}(M)}$; restricting to maximal ideals gives a homeomorphism of maximal spectra. The proof for ${\\mathcal S}^r$ uses an ${\\mathcal S}^r$ version of Lojasiewicz's Nullstellensatz on locally compact $M$, together with a weak version valid for arbitrary $M$, and the bounded case is handled through approximation and localization arguments. A further consequence is that the inclusion ${\\mathcal S}^r(M)\\hookrightarrow{\\mathcal S}^0(M)$ induces an isomorphism of fraction fields $\\kappa(p\\cap{\\mathcal S}^r(M))\\cong\\kappa(p)$, so the residue fields are real closed.","pith_inferences":["If the spectrum homeomorphism is taken as a working tool, one can use ${\\mathcal S}^r$ functions as a full set of test functions for semialgebraic topology, since zero sets and spectral data no longer depend on the chosen differentiability class; a natural testable extension would ask whether the same theorem survives when semialgebraic sets are replaced by definable sets in a general o-minimal st","The single failure of real closedness, namely that sums of radical ideals need not be radical, points toward non-coherence as the underlying phenomenon, so the search for examples where ${\\mathcal S}^\\infty\\neq{\\mathcal N}$ could be guided by zero sets whose ideal generators do not behave coherently at singular points.","A direct computation of $\\operatorname{Spec}{\\mathcal S}^r(M)$ for a non-locally compact, two-dimensional semialgebraic set with a problematic point would provide an explicit check of the inverse formula $q\\mapsto\\sqrt{q{\\mathcal S}^0(M)}$ outside the locally compact case."],"forward_implications":["For every $r\\ge1$, the rings ${\\mathcal S}^r(M)$ and ${\\mathcal S}^{r*}(M)$ are Gelfand rings and have Krull dimension equal to $\\dim(M)$, matching the dimension of the continuous semialgebraic rings.","The maximal spectrum of ${\\mathcal S}^r(M)$ is homeomorphic to the semialgebraic Stone\\,{CC}ech compactification of $M$, so the same compactification serves every differentiability class.","Every quotient of ${\\mathcal S}^r(M)$ by a prime ideal has a real closed field of fractions, and ${\\mathcal S}^0(M)$ is the real closure of ${\\mathcal S}^r(M)$; the only obstruction to ${\\mathcal S}^r(M)$ being a real closed ring is that a sum of two radical ideals need not be radical.","A semialgebraic function of class ${\\mathcal C}^\\infty$ is locally Nash at every point, and on coherent Nash sets the ring ${\\mathcal S}^\\infty(M)$ coincides with the ring of Nash functions ${\\mathcal N}(M)$.","Any spectral property of the continuous semialgebraic rings, such as spectral maps induced by semialgebraic maps, dimension bounds, and Gelfandness, transfers automatically to the differentiable rings ${\\mathcal S}^r(M)$ and ${\\mathcal S}^{r*}(M)$."],"supporting_citations":[{"why":"Supplies the semialgebraic Whitney extension theorem used as Fact 2.4: jets on closed semialgebraic sets extend to ${\\mathcal C}^r$ functions on all of $\\mathbb{R}^m$.","marker":"[Th]"},{"why":"Introduces the jet-based definition of ${\\mathcal S}^r$-functions used throughout the paper.","marker":"[ATh]"},{"why":"Provides the classical Whitney extension theory and the Taylor remainder conditions that the paper adapts to semialgebraic jets.","marker":"[M]"},{"why":"Provides the standard real-algebraic background, including semialgebraic retractions, Nash approximation, and dimension theory, invoked in several reductions.","marker":"[BCR]"},{"why":"Is the source for real closed rings, the real closure construction, and the fact that residue fields of real closed rings are real closed.","marker":"[S4]"},{"why":"Gives the semialgebraic retraction theorem used to reduce locally compact sets to closed sets in Lemma 2.12.","marker":"[DK1]"},{"why":"Supplies the depth and dimension facts for rings of continuous semialgebraic functions that the differentiable case inherits through the spectra homeomorphism.","marker":"[FG6]"},{"why":"Gives the localization correspondence between prime ideals and primes avoiding a multiplicative set, used to transfer the unbounded spectrum result to bounded rings.","marker":"[AM]"}],"fun_headline_variants":["All C^r semialgebraic rings share one Zariski spectrum","Same Zariski spectrum for every C^r semialgebraic ring","C^r and C^0 semialgebraic rings have identical spectra","One spectrum fits all C^r semialgebraic rings"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the semialgebraic Whitney extension theorem: every ${\\mathcal C}^r$ semialgebraic function on a closed semialgebraic set, together with compatible data for all partial derivatives up to order $r$, can be extended to an actual ${\\mathcal C}^r$ semialgebraic function on the whole ambient space; if this extension step fails in even one case, the bridge to Lojasiewicz's Nullstellensatz and hence to the spectra homeomorphism collapses.","fun_headline_variants_meta":{"raw":{"variants":["All C^r semialgebraic rings share one Zariski spectrum","Same Zariski spectrum for every C^r semialgebraic ring","C^r and C^0 semialgebraic rings have identical spectra","One spectrum fits all C^r semialgebraic rings"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0004,"raw_usage":{"total_tokens":2200,"prompt_tokens":1163,"completion_tokens":1037,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":779,"completion_tokens_details":{"reasoning_tokens":966}},"tokens_in":779,"tokens_out":1037,"duration_ms":11005,"temperature":1.0,"reasoning_tokens":966,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:29:04.197602+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a semialgebraic set $M$ and a prime ideal $q$ of ${\\mathcal S}^r(M)$ for which $\\sqrt{q{\\mathcal S}^0(M)}\\cap{\\mathcal S}^r(M)$ strictly contains $q$; the paper's inverse formula says these two sets must be equal, so such a pair would falsify Theorem 1.2 and the real-closure statement.","supporting_citations":[],"review_version":1}