{"id":"afaf95b4-4bc6-433f-8d6e-fc52fa2363a8","arxiv_id":"2508.07168","paper_version":2,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Generalizes symplectic moment map theory to non-degenerate 'momentumly closed' two-forms on almost Hermitian manifolds, claiming Darboux-Weinstein, convexity, reduction, and Kirwan-Ness results.","lead":"This paper introduces 'momentumly closed' two-forms, objects that generalize symplectic forms in almost Hermitian geometry, together with a generalized moment map. It claims that the classical theory of moment maps, including normal forms, convexity, reduction, and stratification, extends to this broader setting.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Momentumly closed condition may be too weak to support Darboux-Weinstein; garbled text prevents verification","rationale":"Good-faith reading: the abstract and section headings indicate a plausible program: introduce 'momentumly closed' 2-forms and prove analogues of the four classical theorems. The provided text is character-garbled; no definition, statement, or proof is inspectable. The reader's weakest_assumption correctly identifies the definition of momentumly closed as load-bearing. I agree. I add specificity: the first point of failure is the Darboux-Weinstein variant. Classical proofs require dω=0 to get a flat model; if momentumly closed does not force local closedness (or include a torsion term that vanishes in the normal coordinates), the remaining theorems inherit the gap. This is an internal sufficiency concern, not a disagreement with consensus. I propose a two-part check: verify the trivial-action reduction to dω=0, and test the construction on the nearly Kähler S^6. Because the text cannot be read, the concern cannot be resolved here; the verdict stays UNVERDICTED. No change to the reader's assessment.","tokens_in":23326,"tokens_out":5589,"duration_ms":53341,"concrete_test":"Obtain a clean copy of arXiv:2508.07168 and isolate Definition (momentumly closed). Do two checks. (1) Consistency: verify that if the group action is trivial and the generalized moment map is constant, the condition is equivalent to dω=0. (2) Non-vacuity: take the standard nearly Kähler structure on S^6 with ω(X,Y)=g(JX,Y) and its almost complex structure; compute whether any choice of moment map makes this ω momentumly closed. If yes, run the paper's Darboux-Weinstein construction near a point where the moment map is regular; if it yields local coordinates with ω constant, compare with the known result dω≠0 on S^6, yielding a contradiction. If no, identify the exact torsion term that fails and check whether the paper's theorem has extra hypotheses that exclude this example.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim generalizes symplectic moment-map theory by replacing closedness of the symplectic form with 'momentumly closed' (Section 2). The load-bearing point is whether this condition is strong enough to yield the normal form behind the Darboux-Weinstein variant that the paper then uses for convexity, reduction, and Kirwan-Ness stratification. In the classical argument, dω=0 is used in three unavoidable places: Moser's trick/path method to get a constant-coefficient model, Hamiltonian vector fields being symplectic (Lie derivative L_X_f ω=0) so that moment-map components Poisson commute, and closedness of the reduced form. If 'momentumly closed' only involves the moment map and the almost complex structure, it may still allow dω≠0 on non-integrable manifolds. Then the Darboux-Weinstein normal form cannot be the standard constant form: any local equivalence to a constant-coefficient model forces dω=0 in that chart. The provided full text is unreadable (mojibake), so the actual definition and proof cannot be checked. The paper itself does not include a readable limitation statement, but the central assumption is precisely the place where the structure could fail.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces the notion of a 'momentumly closed' two-form and an associated generalized moment map in the almost Hermitian setting, and announces four families of results: a Darboux-Weinstein-type normal form, convexity of the generalized moment map, existence of a reduction space, and expected properties of the Kirwan-Ness stratification. The abstract is readable and is standard in shape, but the provided full text is severely corrupted (mojibake); no definition, theorem statement, or proof is legible. Consequently, none of the announced results can be inspected or verified from the submitted text.","tokens_in":23523,"tokens_out":4052,"duration_ms":40532,"significance":"If the announced results are correct, they would constitute a meaningful extension of classical moment-map theory to a non-symplectic almost Hermitian setting, potentially connecting work on almost Kähler geometry, convexity, and reduction. The claimed scope—Darboux-Weinstein, convexity, reduction, and Kirwan-Ness stratification—would be a substantial contribution. However, the significance is entirely conditional: the manuscript does not provide inspectable proofs, and no machine-checked proofs, reproducible code, parameter-free derivations, or falsifiable predictions are visible. This referee therefore cannot currently certify any of the claimed theorems.","major_comments":[{"comment":"The entire body of the manuscript is corrupted and unreadable; it consists of mojibake with only fragments of section headings and occasional formulas. This is load-bearing: the abstract announces four theorem families (Darboux-Weinstein variant, convexity, reduction, Kirwan-Ness stratification), but no definition of 'momentumly closed', no precise theorem statements, and no proofs are accessible. I cannot verify that the central concept is well-defined or that the announced results are proved. A readable version is an absolute prerequisite for further review.","section":"Full text"},{"comment":"The Darboux-Weinstein claim is the load-bearing premise for the remaining results. In the classical argument, dω=0 is used in at least three unavoidable places: the Moser path method, the preservation of ω by Hamiltonian vector fields (so moment-map components Poisson commute), and closedness of the reduced form. The abstract only says that a 'momentumly closed' two-form is considered; it does not state whether this condition implies local closedness of the form. If it does not, the standard constant-coefficient Darboux normal form cannot hold, because any local equivalence to a constant form forces dω=0 in that chart. The manuscript must state precisely what the Darboux-Weinstein normal form is and prove it from the definition; otherwise the convexity, reduction, and Kirwan-Ness arguments have no visible foundation.","section":"Abstract / Section 2"},{"comment":"Even at the level of announced results, the hypotheses are underspecified. Convexity of moment images classically requires properness and a connected group action; Kirwan-Ness stratification requires a real moment map and a GIT-type setup. The abstract does not state these hypotheses, and the unreadable body does not allow me to check them. I would need precise statements with all hypotheses before assessing whether the generalization is sound. This is not a complaint about disagreement with consensus; it is a request for the minimal information required to evaluate the claims.","section":"Abstract (reduction and Kirwan-Ness)"}],"minor_comments":[{"comment":"The phrase 'momentumly closed forms' appears in the first sentence, while later the object is described as a 'non-degenerate momentumly closed two-form.' Please use consistent terminology throughout; if the term is introduced, define it in the abstract or in a clearly marked preliminary section.","section":"Abstract"},{"comment":"The arXiv identifier indicates version 2 (dated 24 Jun 2026), but no list of changes from v1 is provided. If this is a resubmission, please summarize the changes in the cover letter or a revision note.","section":"Title page"},{"comment":"The text is corrupted beyond readability; likely a font/encoding problem in the PDF or source. Please ensure that the manuscript compiles and renders correctly before any resubmission, including all displayed equations and the bibliography.","section":"Full text"},{"comment":"Consider whether 'momentumly closed' is the intended term; 'momentum-closed' or 'moment-map closed' might be more transparent to readers familiar with moment maps.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The submitted file is effectively unreadable: the body is mojibake with only fragments of headings and formulas. I recommend returning the manuscript to the authors to resubmit a properly rendered PDF or source file before technical review. If the corruption is an artifact of the submission pipeline, the editor should request the original file. The technical concern about the Darboux-Weinstein step should be addressed explicitly in the revision; the authors need to state the exact definition of 'momentumly closed' and prove the normal form from that definition. With the current text, no meaningful technical review is possible."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the paper's program is coherent and plausibly useful — replacing dω=0 by a \"momentumly closed\" condition and rebuilding moment map theory in an almost Hermitian setting. That is a real idea worth someone's time. But the full text I received is mojibake; I can read only the abstract and section headings. So I can't certify any of the four claimed theorems.\n\nWhat looks genuinely new: the notion of momentumly closed forms and the generalized moment map, and the plan to push Darboux-Weinstein, convexity, reduction, and Kirwan-Ness stratification through in that setting. The abstract is sensible, and the claims are recognizably in the shape of the classical theory. If the technical work checks out, this would be a useful extension — not a paradigm shift, but a solid toolkit contribution.\n\nSoft spots, in proportion: the load-bearing point is the definition itself. The stress-test note worries that \"momentumly closed\" may be too weak to supply the normal form behind Darboux-Weinstein, since Moser's argument, the symplectic nature of Hamiltonian vector fields, and closedness of the reduced form all use dω=0 in the classical proof. That is a legitimate thing to check, but it is not evidence of an error. The garbled text prevents even checking whether the condition guarantees local constancy of ω or just some weaker compatibility with J and the moment map. So the soft spot is not a flaw in the paper; it's a verification gap created by the artifact I was given.\n\nOn the abstract alone, the paper looks like honest, standard work. No red flags, no fitted parameters, no visible circularity. The citation pattern cannot be assessed because the bibliography is inside the mojibake.\n\nRecommendation: if the actual arXiv PDF is readable, send it to a competent symplectic/almost complex geometer. The referee should focus on Section 2's definition and the Darboux-Weinstein proof first; the other three results likely hang off it. My own verdict is unverified, and I'd want to read a clean copy before citing it. But an editor should not desk-reject on the basis of this corrupted copy.","headline":"Generalized moment maps in almost Hermitian geometry: the abstract promises a coherent extension, but the supplied text is garbled, so the four theorem families remain unverified.","tokens_in":24047,"tokens_out":2757,"would_cite":false,"duration_ms":24625,"reading_group":"maybe","serious_thinker":"unclear","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53D20","53D05","53C15"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that replacing a symplectic form by a non-degenerate 'momentumly closed' two-form, together with a generalized moment map, carries the core of moment-map geometry into the almost Hermitian setting.","keywords":["momentumly closed forms","generalized moment maps","almost Hermitian geometry","Darboux-Weinstein theorem","convexity of moment maps","symplectic reduction","Kirwan-Ness stratification","complex quotients"],"falsifier":"Construct a compact almost Hermitian manifold admitting a non-degenerate momentumly closed form but no symplectic form, compute the image of its generalized moment map, and check convexity. A single example with a non-convex image would refute the convexity theorem; a convex image on a genuinely non-symplectic manifold would demonstrate the generalization is nontrivial.","tokens_in":1347,"feed_emoji":"📐","tokens_out":1589,"duration_ms":67024,"temperature":0.7,"pith_summary":"This paper introduces \"momentumly closed\" two-forms as a generalization of symplectic forms in almost Hermitian geometry, along with a corresponding generalized moment map. It argues that the main theorems of classical moment-map theory—a Darboux-Weinstein normal-form theorem, convexity of the moment-map image, the existence of a reduced space, and the expected Kirwan-Ness stratification—all survive when closedness of the form is replaced by momentum closedness. A sympathetic reader would care because this widens the territory in which moment-map and symplectic-reduction techniques can operate: manifolds that admit no symplectic form may still admit the structure needed for convexity, reduction, and quotient geometry.","feed_headline":"Moment maps exist without symplectic forms","feed_subtitle":"A weaker 'momentumly closed' condition preserves convexity, reduction, and Kirwan-Ness stratification on almost Hermitian manifolds.","key_machinery":"The named central object is the \"momentumly closed form\": a non-degenerate two-form $\\omega$ on an almost Hermitian manifold satisfying a differential condition that replaces $d\\omega = 0$ in this setting. Its partner is the generalized moment map $\\mu$, the analogue of the symplectic moment map. These two objects carry the whole argument: the Darboux-Weinstein normal form, the gradient behavior behind convexity, the reduction construction, and the stratification all are claimed to follow from momentum closedness rather than from the existence of a globally symplectic form.","core_discovery":"The central discovery is the notion of a momentumly closed two-form: a non-degenerate two-form on an almost Hermitian manifold that plays the role classically played by a closed symplectic form. With a generalized moment map attached to it, the paper claims that the following classical results extend to this setting: a variant of the Darboux-Weinstein theorem, convexity of the generalized moment map, construction of a reduction space, and the main properties of the Kirwan-Ness stratification. In other words, the paper asserts that the load-bearing condition in moment-map geometry is not integrability or closedness in the usual sense, but this suitably defined momentum-closed condition.","pith_inferences":["One test this invites is to look for compact almost Hermitian manifolds with no symplectic structure, such as nearly Kähler $S^6$, and check whether a momentumly closed form actually exists there; if it does, the framework has real content beyond the symplectic setting.","The convexity theorem suggests a possible bridge to non-Kähler complex geometry: momentumly closed forms might provide symplectic-type coordinates and convex moment images on complex manifolds that carry no Kähler or symplectic form.","A natural extension the paper leaves implicit is whether momentum closedness is stable under operations such as products and reductions; if so, the class of admissible manifolds would be closed under natural geometric construction, making the framework a working calculus rather than a single theorem.","The Darboux-Weinstein variant hints at a local rigidity statement: any two momentumly closed forms agreeing to first order at a point should be related by a local diffeomorphism, a statement one could verify explicitly in small-dimensional almost Hermitian examples."],"forward_implications":["Almost Hermitian manifolds admitting a non-degenerate momentumly closed form gain local normal-form control analogous to symplectic manifolds.","The generalized moment map has a convex image, extending the classical convexity theorems for moment maps to this broader class.","A reduction space exists for the generalized moment map, so quotient constructions familiar from symplectic reduction can be performed without a symplectic form.","The Kirwan-Ness stratification applies, giving the expected orbit-type structure for gradient flow of the norm-squared generalized moment map.","In the integrable symplectic case, the new notions reduce to classical symplectic forms and moment maps, so the theory is a genuine extension rather than a replacement."],"supporting_citations":[],"fun_headline_variants":["Symplectic not required: momentumly closed forms do the job","Moment map theory survives without integrability","Generalized moment maps: a weaker condition that still works","Beyond symplectic: moment maps on almost Hermitian spaces"],"cache_read_input_tokens":25856,"weakest_assumption_plain":"Everything rests on \"momentumly closed\" being strong enough to play the role of closedness: if this condition does not force the needed Darboux-Weinstein normal form and Hamiltonian-type gradient behavior, the convexity, reduction, and stratification theorems would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Symplectic not required: momentumly closed forms do the job","Moment map theory survives without integrability","Generalized moment maps: a weaker condition that still works","Beyond symplectic: moment maps on almost Hermitian spaces"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001028,"raw_usage":{"total_tokens":4109,"prompt_tokens":624,"completion_tokens":3485,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":368,"completion_tokens_details":{"reasoning_tokens":3418}},"tokens_in":368,"tokens_out":3485,"duration_ms":27689,"temperature":1.0,"reasoning_tokens":3418,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T22:16:40.365708+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a compact almost Hermitian manifold admitting a non-degenerate momentumly closed form but no symplectic form, compute the image of its generalized moment map, and check convexity. A single example with a non-convex image would refute the convexity theorem; a convex image on a genuinely non-symplectic manifold would demonstrate the generalization is nontrivial.","supporting_citations":[],"review_version":1}