{"id":"b5fcd531-492c-43ad-9ee2-dc621639ed17","arxiv_id":"2505.18841","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A self-stress on a framework of surfaces yields a monodromy-free piecewise-linear lifting whenever the self-stress space has dimension greater than 3 times the first Betti number of the surface.","lead":"Mathematicians give a topological recipe for lifting a flat wireframe drawing into a three-dimensional surface, even when the drawing's edges cross. The recipe works for oriented and non-oriented surfaces and includes a dimension test that forces a non-trivial lift.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Thm 3.9's rank count ignores that H1 torsion classes need not be null-homotopic; Klein-bottle monodromy conditions may exceed 3b1(S).","rationale":"The paper's central claim is Theorem 3.9, and its proof is a dimension count: the monodromy-free subspace of self-stresses has codimension at most 3b1(S). The load-bearing step is the assertion that finite-order homology classes impose no conditions. Proposition 3.8 is valid only for face-loops with a null-homotopic multiple in the oriented face-path groupoid, which is a statement about the fundamental group, not about H1. On closed nonorientable surfaces, H1(S;Z) has a Z/2 summand, but its generator need not have finite order in the fundamental group: in the Klein bottle it is the orientation-reversing loop, whose double is null-homologous but not null-homotopic. Therefore the proof's rank count is not established, and the true codimension could be larger than 3b1(S). The reader's weakest_assumption identifies exactly this spot, so I agree with the reader. I did not recommend REJECT because the theorem may be intended for surfaces with nonempty boundary, in which case H1 is torsion-free and the specific objection lapses; a concrete rank computation on an explicit Klein-bottle framework can settle whether the extra conditions are independent. Until that check is run, the conditional verdict is appropriate, so the reader's verdict needs no change.","tokens_in":6927,"tokens_out":23226,"duration_ms":213467,"concrete_test":"Use the 3×2 rectangular grid with Klein-bottle identifications (left/right edges identified with a half-twist, top/bottom identified), giving v=6, e=12, f=6. Insert one diagonal in a face so the 1-skeleton is simple and e=13, f=7; for generic vertex coordinates the self-stress space has dimension d = e-(2v-3) = 4. Let M_1 and M_2 be the affine monodromy linear maps (coefficients of x, y, constant) induced by a face-path representing the H1 generator of the Klein bottle and by the orientation-reversing loop representing the Z/2 homology class, realized via the dual path in the orientation double cover. Stack the two 3×4 matrices and compute the rank over R. If the stacked rank is 6, every nonzero self-stress has nonzero monodromy, contradicting Theorem 3.9 for this framework (d=4 > 3=3b1). If the second block has rank 0, the finite-order dismissal is vindicated for this class.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 3.9 rests on the proof's rank count: 'Every infinite generator of H1(S) provides three linear conditions... We can disregard finite generators due to Proposition 3.8 above. Hence we have precisely 3 b1(S) linear conditions.' The step from Proposition 3.8 to 'disregard finite generators' is invalid for closed nonorientable surfaces. Proposition 3.8 kills a face-loop only when some positive multiple of the loop is null-homotopic as an oriented face-path, i.e., when the loop has finite order in the relevant fundamental groupoid. A generator of the Z/2 summand of H1(S;Z) on a closed nonorientable surface is represented by an orientation-reversing loop whose double is null-homologous but need not be null-homotopic. For the Klein bottle, H1 = Z ⊕ Z/2; the orientation-reversing generator lifts in the orientation double cover (a torus) to a path between the two sheets whose square is a nontrivial loop in that torus. Thus Proposition 3.8 does not apply, and the monodromy of that loop need not vanish. The proof never establishes that monodromy is a homomorphism from H1(S;Z) to the three-dimensional affine group; without that, the asserted count of 3b1(S) conditions is unsupported. The theorem's statement does not exclude closed surfaces, and Remark 3.10's comparison with the closed oriented genus-g theorem suggests closed surfaces are in scope. For surfaces with nonempty boundary the H1 torsion issue disappears, but the proof does not state that restriction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a topological definition of Maxwell-Cremona liftings for frameworks that are 1-skeleta of polygonal surfaces in the plane, allowing non-planar embeddings and non-orientable surfaces. A self-stress defines an elementary lift across an edge, and summing these lifts along an oriented face-path gives an affine function; Theorem 2.8 states that this lift is invariant under homotopy of oriented face-paths. A stress is called monodromy-free when every face-loop has zero lift, and such a stress yields a single-valued piecewise-linear height function on the surface. The main result, Theorem 3.9, claims that if the space of self-stresses has dimension d > 3 b1(S), then there exists a nontrivial monodromy-free polyhedral lifting. The paper also proves a bijection for surfaces with b1(S)=0 and gives a worked example of a framework with one trivial and one nontrivial monodromy.","tokens_in":7178,"tokens_out":12096,"duration_ms":112911,"significance":"If the main estimate is correct, the paper provides an explicit constructive criterion for liftability of non-planar frameworks and recovers the known 6g threshold for oriented surfaces of genus g. The elementary-lift formula is explicit and self-contained, and the homotopy invariance statement is a useful contribution that does not depend on prior work. The example is concrete and illustrates the distinction between trivial and nontrivial monodromy. However, the proof of the central theorem is not complete enough to establish the claimed estimate in full generality, especially for closed nonorientable surfaces.","major_comments":[{"comment":"Theorem 3.9's proof dismisses finite generators of H1(S) by invoking Proposition 3.8, but Proposition 3.8 applies only to face-loops of finite order in the fundamental group(oid), since its proof uses that mγ is null-homotopic. A finite-order class in H1(S;Z) need not have finite order in π1(S): for the Klein bottle, the Z/2 summand of H1(S;Z)=Z⊕Z/2 is represented by an orientation-reversing loop whose double is null-homologous but not null-homotopic, as it lifts to a nontrivial loop in the orientation double cover, a torus. Thus the step 'We can disregard finite generators' is invalid for closed nonorientable surfaces, and the theorem's statement does not exclude them. If Theorem 3.9 is intended only for surfaces with nonempty boundary, where H1(S) is torsion-free, that restriction must be stated and Remark 3.10's comparison with closed oriented surfaces should be qualified.","section":"Section 3.3, Proposition 3.8 and Theorem 3.9"},{"comment":"The sentence 'Every infinite generator of H1(S) provides three linear conditions' asserts without proof that the monodromy of a face-loop depends only on its homology class and that the map from the stress space to the three coefficients (a,b,c) is linear. Homotopy invariance (Theorem 2.8) gives well-definedness on the relevant path groupoid, but the paper does not prove the abelianization step, nor does it state or prove linearity of the monodromy coefficients as functions of the self-stress w. Without these facts, the count of 3b1(S) linear conditions is not established, and the conclusion d > 3b1(S) does not follow from the argument as written.","section":"Section 3.3, proof of Theorem 3.9"},{"comment":"Proposition 3.8 is missing its stated conclusion: the statement says only 'Consider a lift with respect to an oriented face-loop γ that represents a finite order element in the surface S' and then proceeds directly to the proof. The proposition should explicitly conclude that the lift τ_{γ,w}(f,f) is the zero function; as written, the reader must infer the intended claim from the proof.","section":"Section 3.3, Proposition 3.8"}],"minor_comments":[{"comment":"The phrase 'Betty number' should be 'Betti number', and 'the dimension of the space of stresses ... equals to d' should be 'equals d'.","section":"Theorem 3.9"},{"comment":"There are several typographical errors, including 'particulary' and 'surfaces' in the introduction; a careful proofreading pass is needed.","section":"Introduction"},{"comment":"In Definition 2.2, 'a sequence ofF' should read 'a sequence of faces F'; the same definition could also clarify that a face-loop is a face-path with f0 = fN.","section":"Definition 2.2"},{"comment":"The proof of the second elementary move states that the second bracket is zero 'corresponds to the equilibrium condition at the vertex', but the sign bookkeeping that ties the determinant terms to the equilibrium equation is not shown; a short derivation would make the proof easier to check, especially since the paper treats non-oriented surfaces.","section":"Section 2.4"},{"comment":"The proof of Theorem 3.11 is very terse: the sentence 'monodromy along any face-path uniquely defines the lifting' should be expanded to explain how the stresses are recovered from adjacent face heights and why the map is surjective.","section":"Section 3.3, Theorem 3.11"}],"recommendation":"major_revision","confidential_remarks":"The central gap is localized to the proof of Theorem 3.9. The authors can likely repair the issue by either restricting Theorem 3.9 to surfaces with nonempty boundary (where H1 is torsion-free) or by proving that monodromy factors through H1(S;Z) and is linear in the stress. I do not see concerns about novelty or attribution beyond the need to clarify the relationship to [4] in the nonorientable setting."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Karpenkov, Servatius, and Servatius give a topological definition of Maxwell-Cremona liftings for frameworks that are 1-skeleta of polygonal surfaces, allowing non-planar and non-oriented cases. The genuinely new part is the non-oriented version: Theorem 3.9, claiming that if the stress space has dimension d > 3 b1(S), then there is a nontrivial monodromy-free lifting. The oriented case is explicitly acknowledged to coincide with Theorem 45 of Cooperband-Ghrist-Hansen. The paper also gives an explicit homotopy-invariant lift formula along face-paths, which is a useful working tool.\n\nWhat the paper does well: the definitions are clear, the homotopy invariance argument (Theorem 2.8) is careful, and the toric example is concrete and self-contained. The authors honestly compare with the prior cosheaf result rather than overselling novelty.\n\nThe soft spot is the proof of Theorem 3.9. The proof compresses the rank argument into one sentence: each infinite H1 generator gives three linear conditions, and 'finite generators' are disregarded by Proposition 3.8. That step is not justified for closed non-orientable surfaces. Proposition 3.8 applies only to loops with finite order in the fundamental group(oid) of the surface. A generator of the Z/2 summand of H1 on a closed non-orientable surface (e.g., the orientation-reversing loop on the Klein bottle) has infinite order in π1; its double is null-homologous but not null-homotopic. So Proposition 3.8 does not kill it, and the count of 3 b1(S) conditions is unsupported for closed surfaces. If 'with boundary' in the theorem means 'nonempty boundary,' then H1 is torsion-free and the gap is minor—just delete the misleading sentence about finite generators. But the paper never says that, and the comparison in Remark 3.10 with the closed oriented genus case suggests closed surfaces are in scope. Either way the theorem needs a revision that states the boundary hypothesis precisely and proves the rank count, including a treatment of orientation-reversing loops.\n\nThis is not a fatal flaw: the core idea is sound, the definitions are reusable, and the explicit formula is a contribution. A serious referee should ask for the rank argument to be expanded. I would cite this paper for the definition and formula, and would bring it to a reading group with the caveat that Theorem 3.9's proof needs discussion.","headline":"A useful topological extension of Maxwell-Cremona with a genuinely new non-oriented bound, but the proof of the main theorem has a gap around H1 torsion that needs fixing.","tokens_in":7755,"tokens_out":4355,"would_cite":true,"duration_ms":35709,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["52C25","05C10","57M20"],"pacs":[],"model":"deepseek-v4-flash","headline":"If the space of self-stresses of a surface's 1-skeleton has dimension greater than three times the first Betti number, the framework admits a non-trivial monodromy-free polyhedral lifting.","keywords":["self-stress","Maxwell-Cremona lifting","monodromy","polygonal surface","non-planar framework","first Betti number","reciprocal diagrams","rigidity theory"],"falsifier":"Find a compact polygonal surface $S$ in the plane with $d>3b_1(S)$ for which every self-stress has a nonzero monodromy around some face-loop. Since Theorem 3.9 claims such a framework must have a non-trivial monodromy-free lifting, a single such example, produced by a small triangulation with crossing edges and a stress space of dimension $3b_1(S)+1$, would falsify the theorem.","tokens_in":6688,"feed_emoji":"📐","tokens_out":6220,"duration_ms":45237,"temperature":0.7,"pith_summary":"Classically, Maxwell–Cremona liftings turn a planar framework with a self-stress into a three-dimensional polyhedral surface that projects back to the drawing. This paper gives a topological definition of such liftings that works even when the framework is not planar and the surface is not oriented, and it proves a quantitative guarantee: whenever the self-stress space is larger than three times the first Betti number of the surface, a non-trivial single-valued lifting exists. The construction follows a face-path through the surface, accumulating a piecewise-linear height function whose value is independent of the chosen path up to homotopy. The result matters because it extends a classical correspondence between stresses and spatial liftings to settings where edges may cross, which is relevant to rigidity theory, discrete geometry, and polyhedral combinatorics.","feed_headline":"Stresses above 3 per hole force a surface framework to lift","feed_subtitle":"A topological Maxwell–Cremona correspondence now covers non-planar, non-oriented surface frameworks.","key_machinery":"The central object is the lift along an oriented face-path, assembled from elementary lifts $L_{f_i}^{f_{i-1}}(p)=\\det(p_1-q_1,p_1-p)\\,w(p_iq_i)\\,\\rho_{\\mathrm{discr}}(f_0,f_1)$, where $w(p_iq_i)$ is the stress on the shared edge and $\\rho_{\\mathrm{discr}}$ merely records whether two faces are distinct. This sum is homotopy-invariant because it is unchanged by the two elementary moves: inserting or deleting a doubled face at an edge, and inserting or deleting a loop of faces around a vertex, the latter vanishing by the equilibrium condition on the stress. The monodromy-free condition—that every face-loop lifts to zero—then lets the path-wise lift be written as $\\tau_w(f,f')$, and the function $\\tau_{w,f}(p)=\\tau_w(f,f_p)(p)$ is the lifting. The dimension bound in Theorem 3.9 comes from counting, for each of the $b_1(S)$ independent homology loops, the three linear conditions that monodromy-freeness imposes on the stress space, with finite-order loops eliminated by Proposition 3.8.","core_discovery":"On the paper's own terms, the central discovery is a topological version of the Maxwell–Cremona correspondence for arbitrary polygonal surfaces in the plane, oriented or not. A self-stress on the 1-skeleton defines, for any oriented face-path, a lift obtained by summing elementary determinants attached to crossed edges; Theorem 2.8 shows this lift depends only on the homotopy class of the path. A stress is monodromy-free precisely when every face-loop lifts to zero, and then it defines a genuine piecewise-linear function on the surface. The main result, Theorem 3.9, states that if the dimension of the space of self-stresses exceeds $3b_1(S)$, where $b_1(S)$ is the first Betti number, then a non-trivial monodromy-free lifting exists; for oriented surfaces this recovers the previously known genus bound, while for non-oriented surfaces it is new. In the simply connected case the correspondence is a bijection between all self-stresses and all liftings that keep one face in the plane $z=0$.","pith_inferences":["The $3b_1(S)$ bound is likely not sharp in general: the three conditions per loop treat the affine group's degrees of freedom, but geometric degeneracies of a particular framework may reduce the number of independent conditions, so the threshold could be lowered for special configurations.","The construction suggests a direct algorithm for producing liftings: compute the stress space, solve the linear monodromy equations, and assemble the height function face by face; this could be implemented for toroidal or other non-planar frameworks arising in graphic statics.","Extending the definition to higher-dimensional complexes along the lines of the authors' earlier work might yield analogous Betti-number bounds for liftings of $d$-dimensional frameworks, with the constant 3 replaced by the dimension of an affine group acting on $\\mathbb{R}^d$.","A natural test case is to search computationally for polygonal surfaces with $b_1=1$ and stress dimension exactly 3 in which every stress has nonzero monodromy; finding such an example would show the constant 3 cannot be improved without additional hypotheses."],"forward_implications":["For oriented surfaces, the theorem reproduces the existing bound: stress dimension above $6g$ guarantees a lifting, since $b_1=2g$.","For non-oriented surfaces, the bound $d>3b_1(S)$ is the first general estimate of this type; it covers surfaces such as a Möbius band or a Klein bottle with boundary.","When $b_1(S)=0$, the correspondence is a bijection: every self-stress gives a unique lifting with a chosen face horizontal, so polyhedra in space are enumerated by their planar projections with self-stresses.","Cutting a surface along edges to remove the infinite generators of $H_1(S)$ produces a fundamental domain, and the lifting is uniquely reconstructed from it as a possibly multivalued covering of the original.","The worked example of the triangular prism shows that a single framework can admit liftings with different monodromy types, including one trivial and one non-trivial monodromy."],"supporting_citations":[{"why":"Supplies the oriented-surface theorem (Theorem 45 there) that Theorem 3.9 extends, as well as the cosheaf framework being replaced by the combinatorial face-path approach.","marker":"[4]"},{"why":"Introduces Maxwell–Cremona liftings and reciprocal figures, the classical object being generalized to non-planar frameworks.","marker":"[12]"},{"why":"Establishes the classical correspondence between equilibrium stresses and liftings of projected polyhedra, the foundation the topological definition builds on.","marker":"[16]"},{"why":"Elaborates the plane stress–lifting correspondence whose geometric content the new definition carries over.","marker":"[5]"}],"fun_headline_variants":["Surface frameworks lift when self-stresses exceed 3 per hole","New bound: >3 self-stresses per hole forces a lifting","Topological Maxwell–Cremona: lift when stress beats 3b1","Monodromy-free liftings guaranteed when stress count > 3b1","Stress count above 3 per Betti number yields a lift"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that each independent loop in the surface contributes exactly three linear conditions on the space of self-stresses, with no extra hidden conditions and with finite-order loops contributing none; if that rank count fails, the dimension bound $d>3b_1(S)$ would not guarantee a lifting.","fun_headline_variants_meta":{"raw":{"variants":["Surface frameworks lift when self-stresses exceed 3 per hole","New bound: >3 self-stresses per hole forces a lifting","Topological Maxwell–Cremona: lift when stress beats 3b1","Monodromy-free liftings guaranteed when stress count > 3b1","Stress count above 3 per Betti number yields a lift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000624,"raw_usage":{"total_tokens":2809,"prompt_tokens":788,"completion_tokens":2021,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":404,"completion_tokens_details":{"reasoning_tokens":1926}},"tokens_in":404,"tokens_out":2021,"duration_ms":11918,"temperature":1.0,"reasoning_tokens":1926,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:25:27.206447+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a compact polygonal surface $S$ in the plane with $d>3b_1(S)$ for which every self-stress has a nonzero monodromy around some face-loop. Since Theorem 3.9 claims such a framework must have a non-trivial monodromy-free lifting, a single such example, produced by a small triangulation with crossing edges and a stress space of dimension $3b_1(S)+1$, would falsify the theorem.","supporting_citations":[{"cited_title":"Clerk Maxwell","cited_arxiv_id":null,"evidence_quote":"Introduces Maxwell–Cremona liftings and reciprocal figures, the classical object being generalized to non-planar frameworks."},{"cited_title":"Motions and stresses of projected polyhedra","cited_arxiv_id":null,"evidence_quote":"Establishes the classical correspondence between equilibrium stresses and liftings of projected polyhedra, the foundation the topological definition builds on."},{"cited_title":"Autocontraintes planes et poly` edres pro- jet´ es","cited_arxiv_id":null,"evidence_quote":"Elaborates the plane stress–lifting correspondence whose geometric content the new definition carries over."}],"review_version":1}