{"id":"13dcbcd7-2da8-4c63-84bf-e867073e5606","arxiv_id":"2412.18944","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A function with only Morse extrema and critical circles on a cylinder, torus, disk, or sphere always factors as a smooth value-reparametrization of the standard height, radius, or projection function, up to a diffeomorphism.","lead":"This paper proves a normal form for smooth circle-valued functions with a special type of degenerate singularity on a cylinder, torus, disk, or sphere. Every such function is shown to equal a value-reparametrization of the simplest Morse function after a change of coordinates.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sphere proof in §10.3 misidentifies U1 = Q∩V1 as a disk; with critical circles it is an annulus and f=g is only available on the empty set V0∩U1, so Corollary 10.2 cannot be invoked.","rationale":"The reader's conditional verdict is reasonable, but the weakest spot is not the missing S2 case of Proposition 3.6. Section 10.3 contains a concrete internal inconsistency: U1 is an annulus when critical circles exist, yet Corollary 10.2(1) is a disk theorem, and the equality f=g is claimed on the empty intersection V0∩U1. The gap is repairable by treating the cylinder Q directly with Lemma 9.5 and then capping with the polar disks, but that repair is not present in the manuscript. Since the theorem is nevertheless plausible and the surrounding lemmas are standard, I do not recommend rejection; the proof should be completed or the S2 case reproved before the normal form is accepted. Agreement with the reader is partial: we agree there are load-bearing gaps in the S2-related arguments, but disagree that the S2 case of Proposition 3.6 is the decisive one.","tokens_in":18822,"tokens_out":17004,"duration_ms":161732,"concrete_test":"Take the model example f:S2→R with f(θ,t)=t^3 after writing S2 as S1×[-1,1] with each end collapsed to a point; this lies in F^0, has two Morse extrema and one non-extremal critical circle at t=0. Execute the proof of §10.3 for this f: choose V0,V1 as the two polar disks and Q as the cylinder around t=0; compute U1=Q∩V1 and check directly that it is an annulus, not a disk, and that V0∩U1=∅. Then attempt to apply Corollary 10.2; if the step cannot be made valid, provide a corrected argument on the cylinder Q via Lemma 9.5, or show why none exists. This check decides whether the S2 case of Theorem 1.8 is proved or merely asserted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 10.3 is the only proof of Theorem 1.8 for S2. It chooses neighborhoods V_i,W_i of the two extrema z_i and an f-foliated cylinder Q containing all critical circles, with Q∩W_i ≠ ∅, and sets U1=Q∩V1. The text then asserts U1 is a disk and applies Corollary 10.2(1), a disk statement. But if f has any critical circle, that circle separates S2 into two disks containing z0 and z1, so V0 and V1 are disjoint; Q is an annulus and U1 is a collar, hence an annulus, not a disk. Moreover the hypothesis of Corollary 10.2(2) used in Step 3 is stated as 'f=g on V0∩U1', which is empty in this situation. Thus the gluing construction of h and κ for the sphere case is unsupported. A separate gap exists in Proposition 3.6 for S2 (§7.4, left to the reader), but the S2 proof of Theorem 1.8 does not invoke an H-field on S2; it uses only the cylinder/disk reductions, so the annulus-or-disk issue is the more direct threat to the main claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a class F^0(M,P) of smooth functions on oriented surfaces whose critical set consists only of isolated Morse extrema and critical circles modeled on ±y^n, with locally constant boundary behavior. It states that such functions can exist only on a cylinder, torus, disk, or sphere, and proves a normal form theorem: every such f can be written as f = κ ∘ f0 ∘ h^{-1}, where f0 is one of four explicit 'prime' functions, h is a diffeomorphism, and κ is a smooth reparameterization with non-flat critical points away from the image of the critical set and boundary. The proof strategy is to construct an 'H-field' tangent to level sets with only isolated singularities, use its flow to produce a semi-free circle action, then conjugate that action to a standard one and read off κ from a transversal. The cylinder and torus cases are handled directly; the disk and sphere cases are approached through Morse approximation and gluing arguments.","tokens_in":19037,"tokens_out":11529,"duration_ms":112765,"significance":"The intended theorem is substantial: if correct, it gives a one-dimensional reduction of a whole class of functions with non-isolated singularities to four explicit trivial models, in the spirit of Whitney's normal form for even functions. The cylinder and torus proof is explicit, constructive, and free of fitted parameters, and the overall classification claim is falsifiable and likely useful for subsequent work on stabilizers and homotopy types. However, the sphere case and several auxiliary lemmas are not fully proved as written. The main geometric idea is plausible, but the manuscript currently does not establish the full theorem.","major_comments":[{"comment":"The assertion that U1 = Q ∩ V1 is a disk is incorrect when f has critical circles. Since Q is an open f-foliated cylinder containing all critical circles and intersecting both W0 and W1, while V1 is a disk neighborhood of the extremum z1 containing no critical circles, their intersection is an annulus (a collar of the boundary of V1) rather than a disk. Corollary 10.2(1) is a statement about functions on disks and therefore cannot be applied to obtain the diffeomorphism h1 and the reparameterization α1. If f has no critical circles, the theorem is already covered by Lemma 9.2, but the proof as written does not separate that case, and for the critical-circle case this step is invalid.","section":"§10.3, Step 2"},{"comment":"Step 3 invokes Corollary 10.2(2) with the hypothesis that f = g on V0 ∩ U1, calling this set a connected f-foliated neighborhood of ∂U1. However, Lemma 9.9 only guarantees f = g on V0, and V0 ∩ U1 is empty whenever a critical circle separates z0 and z1, which is precisely the situation in which Step 2 is needed. Thus the extension of h1 and α1 by the identity on the complement is not justified. The proof of Theorem 1.8 for S2 therefore does not construct the required diffeomorphism h and reparameterization κ.","section":"§10.3, Step 3"},{"comment":"Proposition 3.6 is stated for all surfaces in the class, but the proof for S2 is omitted with the sentence 'We left details to the reader.' This proposition is the basis for the H-field, and hence for the normalized flows and semi-free circle actions used in Section 5 and in the cylinder/torus part of the proof of Theorem 1.8. A load-bearing lemma of this kind cannot be discharged by a 'similar to the disk case' remark; the sphere case needs an explicit proof or a precise reduction to the disk case.","section":"§7.4"},{"comment":"The inference 'Since f = g on W, it follows that φ_a = ψ_a on W' is not justified. The normalized H-fields in equation (8) are not canonical: Proposition 3.6 produces one vector field among many, and normalizing the period to 1 does not fix the orientation on a leaf. To apply Lemma 5.2, the two actions must coincide on W, and the proof must explain how the H-fields for f and g are chosen compatibly on W, or why any two normalized H-fields agree up to a global rotation on each orbit. Without this, Lemma 9.5 is not proved, and it is used in Lemma 9.7 and Corollary 10.2(2).","section":"Lemma 9.5, first paragraph"},{"comment":"Lemma 9.9 is a load-bearing approximation statement for both the disk and sphere proofs, but its proof is only a sketch. The intermediate function g0 obtained by perturbing f near the critical circles is allowed to have saddles, and the cancellation step must show that all such saddles can be canceled while preserving f = g on V and without creating new critical points outside V. The references to [14] are not enough by themselves; a complete argument, or a precise citation covering this exact situation, is needed.","section":"Lemma 9.9"}],"minor_comments":[{"comment":"There are numerous typographical and OCR artifacts, for example 'smoo th', 'surf aces', and 'represe nts'; the paper needs a careful copyedit.","section":"Abstract and Introduction"},{"comment":"Condition (B) refers to 'f0(∂M)', but f0 is defined on M0, not on M; this should be 'f0(∂M0)'.","section":"Theorem 1.8, condition (B)"},{"comment":"The sentence 'The flow F of a normalized H-field F satisfies F(x,t)=x for each x∈M' should read F(x,1)=x, or equivalently F_{t+1}=F_t; otherwise the flow would be trivial.","section":"Section 5, after equation (8)"},{"comment":"The statement introduces 'a smooth function g : Im(g) → R', but g is already the Morse function; the new reparameterization should be denoted by a different letter such as κ or α.","section":"Lemma 9.7"},{"comment":"Notation such as 'α̃ = α on M\\W' and 'α̃1 = α1 on g(U1\\V0)' mixes subsets of M with subsets of Im(g); the domains and codomains should be made consistent.","section":"Corollary 10.2(2) and §10.3"}],"recommendation":"major_revision","confidential_remarks":"The central theorem is plausible and the cylinder/torus part is solid, but the sphere proof has a concrete geometric error: U1 is an annulus, not a disk, and the gluing hypothesis V0 ∩ U1 can be empty. Several other load-bearing statements are deferred or sketched. I do not think the result is false, but the manuscript needs substantial reworking of the sphere case and the auxiliary lemmas before it can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper proves a normal form for a restricted class of functions (F^0: no saddles, only Morse extrema and critical circles) on four surfaces: f = κ∘f0∘h^{-1}. That is new and useful; the no-critical-circle case reduces to known results, but the critical-circle case is a real extension. The construction of H-fields and the use of free circle actions to get the conjugating diffeomorphism is a nice idea, and the cylinder/torus proof is explicit and credible. The disk case is plausible, resting on approximation and gluing lemmas. There are no fitted parameters, no circular reasoning; the Lemma 9.2 and 9.9 proofs are sketches, but they are standard results with references. I would give all that its due.\n\nThe problem is §10.3, the sphere case—the only proof of Theorem 1.8 for S2. The argument takes Q to be an f-foliated cylinder containing all critical circles and intersecting the two extremal neighborhoods, then sets U1 = Q∩V1 and claims U1 is a disk. It is not: if f has at least one critical circle, that circle separates S2 into two disks, so V0 and V1 are disjoint, and Q is an annulus; U1 is a collar in V1, hence an annulus. Corollary 10.2(1), a disk statement, cannot be applied. Step 3 also uses 'f=g on V0∩U1', but V0∩U1 is empty in this situation, so Corollary 10.2(2) is vacuous. The gluing construction of h and κ on the sphere is therefore unsupported. This is not a small gap; it is the load-bearing step for S2. The proof might be repairable—likely by working with the cylinder Q and Lemma 9.5 directly—but it is not in the paper. Separately, Proposition 3.6 for S2 is left to the reader (§7.4), but the sphere proof does not depend on it, so that is a minor issue.\n\nSo: the central claim is probably true, but as written the main theorem is proven for cylinder, torus, and (modulo sketches) disk, not for sphere. The paper deserves peer review because the result is significant for the classification program and the gap is identifiable and probably fixable. A serious referee should require a corrected S2 proof before acceptance. I would read a revised version.","headline":"New normal form for F^0 on four surfaces; cylinder/torus/disk cases look right, but the sphere proof in §10.3 has a real annulus-vs-disk gap.","tokens_in":19603,"tokens_out":5239,"would_cite":true,"duration_ms":46311,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57R45","57R70","57S25"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that every function in a broad class on four surfaces is a reparameterization of one of four standard functions, up to diffeomorphism.","keywords":["circle-valued functions","critical circles","normal forms","semi-free circle actions","Morse-Bott functions","surfaces","H-fields","degenerate singularities"],"falsifier":"Work out the sphere case of Proposition 3.6 explicitly: take $f\\in F^\\circ(S^2,\\mathbb{R})$ with two isolated extrema and at least one critical circle, write the Hamiltonian field near a non-extremal critical circle where the flow direction reverses, and check whether the three-step correction of Section 7, which is proved only for the cylinder, can be carried out; a single such function for which no level-set-tangent vector field with isolated zeros exists would disprove the normal form.","tokens_in":18567,"feed_emoji":"🔄","tokens_out":7062,"duration_ms":65193,"temperature":0.7,"pith_summary":"This paper aims to prove that an entire class of smooth functions with degenerate singularities on four surfaces can be reduced to a simple normal form. The class $F^\\circ(M,P)$ consists of real- or circle-valued functions on a cylinder, torus, disk, or sphere whose only isolated singularities are Morse extrema and whose non-isolated singularities are critical circles locally of the form $\\pm y^n$. The main theorem claims that every such $f$ decomposes as $f = \\kappa \\circ f_0 \\circ h^{-1}$, where $f_0$ is one of four explicit prime functions, $h$ is a diffeomorphism, and $\\kappa$ is a smooth map of the value interval with only finitely many non-flat critical points, none at the special values. If true, the whole class is described by a single one-dimensional reparameterization of four standard functions, so all remaining complexity sits in $\\kappa$ and in the choice of $h$.","feed_headline":"One reparameterization describes all degenerate functions on four surfaces","feed_subtitle":"Every function in the class breaks into a diffeomorphism plus a value change of a standard height, radius, or projection.","key_machinery":"The load-bearing tool is the H-field (Proposition 3.6): for every $f\\in F^\\circ$, one can find a vector field tangent to the level sets of $f$ whose only singularities are isolated, located at the isolated Morse extrema of $f$. After multiplying by a trajectory-constant factor, its flow has period one and therefore defines a smooth semi-free $S^1$-action on $M$. The proof of Theorem 1.8 uses the classification of free $S^1$-actions to conjugate this action to the standard rotation action, which yields the diffeomorphism $h$; $\\kappa$ is then read off by evaluating $f$ on a curve transversal to the foliation. Two auxiliary devices complete the argument: a lemma that corrects a conjugating diffeomorphism on a cylinder by shifting along flow lines, and an approximation lemma that replaces $f$ near its extremum by a Morse function agreeing with $f$ on a neighborhood, so the disk and sphere cases reduce to the cylinder case.","core_discovery":"The central claim is Theorem 1.8: for any oriented surface $M$ among $S^1\\times[0,1]$, $D^2$, $S^2$, and $T^2$, and any $f\\in F^\\circ(M,P)$, there is a diffeomorphism $h\\colon M_0\\to M$ from the model surface and a smooth $\\kappa\\colon f_0(M_0)\\to P$ such that $f=\\kappa\\circ f_0\\circ h^{-1}$. Here $f_0$ is the prime function: height on the cylinder, $x^2+y^2$ on the disk, the $z$-coordinate on the sphere, or the second circle factor on the torus. The reparameterization $\\kappa$ creates any critical circles of $f$: critical points of $\\kappa$ correspond to critical circles of $f$, and $\\kappa$ must be non-flat at those points and have no critical points at $f_0(\\Sigma_{f_0})$ or $f_0(\\partial M)$. If $f$ has no critical circles, then $\\kappa$ is a diffeomorphism and $f$ is smoothly equivalent to $f_0$. The factorization is not unique because it depends on the choice of $h$, but the existence of such a factorization is what the paper establishes.","pith_inferences":["If the normal form holds, then two functions in the class with the same ordered list of extremal and non-extremal critical circles should differ only by a diffeomorphism and a reparameterization $\\kappa$; this is a testable classification statement the author does not spell out.","The unproved sphere case of Proposition 3.6 is the natural place to look for a counterexample: a sphere function whose critical circles force the Hamiltonian flow to change orientation through a non-extremal circle would need an H-field not constructed by the paper's cylinder-style induction.","The same argument suggests a broader principle: degenerate singularities organized in circles carry no new topological information beyond the order and types of critical values, so normal forms of this kind may extend to other equivariant settings where a semifree circle action exists.","One could test the sharpness of the non-flatness condition on $\\kappa$ by examining a function where the natural $\\kappa$ would have a flat critical point; the theorem predicts such a function actually lies outside $F^\\circ$, which is a computable check."],"forward_implications":["On each of the four surfaces, any function in the class can be pulled back to one of four explicit prime functions by a change of values in the target; classification reduces to classifying smooth maps $\\kappa$ of an interval or circle.","All critical circles of $f$ appear exactly at critical values of $\\kappa$, so the critical set of $f$ is always a linearly ordered (on cylinder, disk, sphere) or cyclically ordered (on torus) family of parallel circles.","Functions without critical circles are smoothly equivalent to the corresponding prime function, recovering the known Morse case as a corollary.","The semi-free $S^1$-action produced by the H-field lives on exactly the four surfaces with nonnegative Euler characteristic, matching the surfaces that can carry such actions; the identity component of the right stabilizer is $S^1$ precisely in this class.","The normal form gives an explicit analytic description of the class $F^\\circ(M,P)$, complementing the earlier homotopy-type result for stabilizers."],"supporting_citations":[{"why":"Supplies the Hamiltonian-like vector field lemma for Morse functions that is the starting point for constructing H-fields.","marker":"[9]"},{"why":"Provides reparameterization of vector fields and period functions, used to normalize the H-field and obtain the semi-free $S^1$-action.","marker":"[10]"},{"why":"Gives the classification of free $S^1$-actions on the cylinder and torus that produces the conjugating diffeomorphism $h$.","marker":"[15]"},{"why":"Supplies smooth shift maps along flows, used in Lemma 5.2 to modify the conjugating diffeomorphism.","marker":"[8]"},{"why":"Provides the Morse-theoretic tools for extending diffeomorphisms and canceling critical points used in the disk and sphere arguments.","marker":"[14]"},{"why":"Earlier result on the homotopy type of stabilizers that motivates the class $F^\\circ$ and Theorem 1.6.","marker":"[2]"},{"why":"The classical even-function decomposition $f(x)=\\alpha(x^2)$ that serves as the conceptual template for the normal form $\\kappa\\circ f_0$.","marker":"[19]"},{"why":"Used to smoothly extend the period function to the whole surface when normalizing the H-field.","marker":"[6]"}],"fun_headline_variants":["One value twist turns any degenerate circle function into a standard one","All four surfaces share one normal form for degenerate circle functions","Reparameterization unifies all degenerate circle functions on four surfaces","Diffeomorphism plus value change: the universal normal form"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that every function in the class, including on the sphere, admits a normalized H-field: a vector field tangent to level sets with only isolated zeros whose flow is a periodic rotation; the sphere case is stated without proof, so the whole normal form stands on that unverified step.","fun_headline_variants_meta":{"raw":{"variants":["One value twist turns any degenerate circle function into a standard one","All four surfaces share one normal form for degenerate circle functions","Reparameterization unifies all degenerate circle functions on four surfaces","Diffeomorphism plus value change: the universal normal form"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00128,"raw_usage":{"total_tokens":5236,"prompt_tokens":952,"completion_tokens":4284,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":568,"completion_tokens_details":{"reasoning_tokens":4214}},"tokens_in":568,"tokens_out":4284,"duration_ms":27964,"temperature":1.0,"reasoning_tokens":4214,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:18:47.979465+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Work out the sphere case of Proposition 3.6 explicitly: take $f\\in F^\\circ(S^2,\\mathbb{R})$ with two isolated extrema and at least one critical circle, write the Hamiltonian field near a non-extremal critical circle where the flow direction reverses, and check whether the three-step correction of Section 7, which is proved only for the cylinder, can be carried out; a single such function for which no level-set-tangent vector field with isolated zeros exists would disprove the normal form.","supporting_citations":[{"cited_title":"Homotopy types of stabilizers and or bits of Morse functions on surfaces","cited_arxiv_id":null,"evidence_quote":"Supplies the Hamiltonian-like vector field lemma for Morse functions that is the starting point for constructing H-fields."},{"cited_title":"Reparametrizations of vector fields and their shift maps","cited_arxiv_id":"0907.0354","evidence_quote":"Provides reparameterization of vector fields and period functions, used to normalize the H-field and obtain the semi-free $S^1$-action."},{"cited_title":"Geometry of differential forms , volume 201 of Translations of Mathematical Monographs","cited_arxiv_id":null,"evidence_quote":"Gives the classification of free $S^1$-actions on the cylinder and torus that produces the conjugating diffeomorphism $h$."},{"cited_title":"Smooth shifts along trajectories of flows","cited_arxiv_id":null,"evidence_quote":"Supplies smooth shift maps along flows, used in Lemma 5.2 to modify the conjugating diffeomorphism."},{"cited_title":"An introduction to Morse theory , volume 208 of Translations of Mathematical Mono- graphs","cited_arxiv_id":null,"evidence_quote":"Provides the Morse-theoretic tools for extending diffeomorphisms and canceling critical points used in the disk and sphere arguments."},{"cited_title":"Homotopy type of stabilizers of smooth functions with non-isolated singularities on surfaces","cited_arxiv_id":"2305.08255","evidence_quote":"Earlier result on the homotopy type of stabilizers that motivates the class $F^\\circ$ and Theorem 1.6."},{"cited_title":"Differentiable even functions","cited_arxiv_id":null,"evidence_quote":"The classical even-function decomposition $f(x)=\\alpha(x^2)$ that serves as the conceptual template for the normal form $\\kappa\\circ f_0$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Used to smoothly extend the period function to the whole surface when normalizing the H-field."}],"review_version":1}