{"id":"77128575-7ac5-4a34-9656-976e86c06416","arxiv_id":"2501.13627","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For any open, fiberwise dense first-order differential relation, every piecewise smooth section can be epsilon-jiggled to a solution, and the space of piecewise solutions is weakly equivalent to the space of continuous sections.","lead":"A new jiggling theorem lets any section of a fiber bundle be perturbed, after subdividing the domain, into a piecewise smooth solution of any open and fiberwise dense first-order differential relation. The result is an h-principle without formal homotopy data that also works on closed manifolds.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 7.4 rests on an unproved relative piecewise-smooth approximation/extension step; if that step fails, the claimed weak equivalence does not follow.","rationale":"The reader’s formal weakest assumption is the fiberwise-density condition, which I agree is necessary and natural: it is what makes local tilting possible via Lemma 5.1 and Corollary 5.5. However, the real obstacle to accepting Theorem 7.4 is the unproved relative approximation/extension step in its proof, exactly as the reader’s rationale notes. I do not see circularity, parameter-fitting, or an invented entity. The quantitative core of the paper — crystalline subdivisions, coloring, slope perturbations, and the estimates in Lemmas 4.2, 4.6, 5.6, 5.8, and 6.1 — is substantial and appears internally coherent. The load-bearing gap is narrower: to pass from continuous sections to piecewise smooth sections relative to a boundary that is already a piecewise smooth solution, one needs a lemma that is standard but nontrivial, especially for a bundle over a non-compact manifold and for simplices of dimension n ≥ 2 with corners. The same issue recurs in Lemma 7.3, where filling a horn requires extending a piecewise smooth solution on the horn to all of M × ∆n. The paper’s jiggling theorem cannot be applied until such an extension is produced. Because this is a missing justification rather than a demonstrated contradiction, I agree with the reader’s CONDITIONAL verdict; my concern does not change that verdict. The proposed concrete test — writing out the extension lemma in the minimal case X = M × ∆^1, A = M × ∂∆^1 with arbitrary bundles and non-compact M — would settle whether the missing step is routine or genuine.","tokens_in":33648,"tokens_out":11583,"duration_ms":116511,"concrete_test":"Isolate the step as a lemma: given a triangulated manifold X, a subcomplex A with a collar, a smooth fiber bundle E → X, an open relation R ⊂ J^1(E), and a continuous section F whose restriction to A is piecewise smooth and satisfies j^1F ∈ R on Op(A), construct a piecewise smooth F′ with F′|A = F|A, F′ C^0-close to F, and j^1F′ ∈ R on some collar of A. Check this first in the minimal nontrivial case X = M × ∆^1, A = M × ∂∆^1, with M possibly non-compact and E an arbitrary smooth bundle; write the local chart-by-chart construction and verify that the corner conditions at M × {0,1} can be met exactly while keeping the jet in R on a collar. If the lemma can be proved in this case, Theorem 7.4’s proof can be completed; if not, the weak equivalence has a missing hypothesis or is false.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim Theorem 7.4 is proved by two steps: approximate an arbitrary continuous section F by a piecewise smooth section F′ that agrees with F on M × ∂∆n and solves R on Op(M × ∂∆n), then apply jiggling relative to the boundary. The first step is asserted without proof: 'Firstly, we approximate F by a section F′ that is piecewise smooth such that F′ is a solution on Op(M × ∂∆n) and such that F′ agrees with F on M × ∂∆n.' The same extension problem appears in Lemma 7.3, where a piecewise smooth solution on a horn Λ[n,k] must be extended to a piecewise smooth section over M × ∆n that is a solution on a neighborhood of the horn. This is not automatic: the boundary data force the values of F′ on M × ∂∆n exactly, while the jet of F′ on a collar must lie in the open relation R; for n > 1 the collar has corners, so the extension has to be assembled simplex by simplex and across faces. The standard machine of relative smooth approximation plus openness of R on a compact collar probably supplies such a lemma, but it is not written down, and Theorem 6.3 cannot be invoked until F′ exists. The fiberwise-density assumption is not the suspect ingredient here; this missing step depends only on openness of R and on piecewise-smooth approximation relative to a closed subcomplex. Because both Lemma 7.3 and Theorem 7.4 depend on this extension, the h-principle statement is conditional on it.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a generalized jiggling theorem (Theorem 6.3): for any open and fiberwise dense differential relation R ⊂ J^1(E) of first order, every piecewise smooth section of E can be subdivided and C^1-approximated by a piecewise smooth solution of R, with relative and non-compact variants. The proof works by choosing a fine crystalline subdivision, linearizing the section, and then slope-perturbing it color by color, using quantitative estimates on the geometry of crystalline subdivisions. The authors then package this into an h-principle without homotopical assumptions: the simplicial set sSol_PS(R) of piecewise smooth solutions over M × Δ^n is weakly equivalent to the singular complex of the space of continuous sections of E (Theorem 7.4). Applications are given to piecewise transversality, immersions and submersions, contact forms, and a new proof of Thurston's jiggling lemma, including non-compact manifolds.","tokens_in":33886,"tokens_out":12365,"duration_ms":123769,"significance":"If Theorem 6.3 and Theorem 7.4 are correct, this is a substantial contribution to the flexibility of piecewise smooth first-order differential relations. The paper gives a self-contained, quantitative proof of the core jiggling result, with explicit bounds coming from model simplices (Lemma 3.8), linearization (Lemma 4.2), interpolation and join estimates (Lemmas 4.4 and 4.6), and slope-perturbation estimates (Lemma 5.8). The color-by-color induction in Theorem 6.1 is a genuine simplification of Thurston's original argument and is carefully set up. The applications to contact forms and to general position are attractive and show the power of the result. There is no circularity or fitting of parameters in the main construction. The central weakness is in Section 7: the passage from jiggling to the simplicial h-principle contains an unproved relative piecewise-smooth extension/approximation step. This step is load-bearing for both Lemma 7.3 and Theorem 7.4, so the paper needs a dedicated argument there before the announced h-principle is fully established.","major_comments":[{"comment":"The proof of Lemma 7.3 states: 'Extend s to a piecewise smooth section tilde{s} of E over M × Δ^n such that it is piecewise smooth with respect to a triangulation tilde{T} that extends T. We assume that tilde{s} is a solution on Op(M × Λ^n_k).' The second sentence is not automatic from openness of R. A solution on the horn fixes the values and derivatives on the horn, but it gives no control on the normal derivatives used to extend into the missing faces; for n ≥ 2 the collars of the horn meet in corners, so the extension has to be assembled simplex by simplex and across faces. Because the Kan property of sSol_PS(R) is asserted from this lemma and is then used in the weak-equivalence criterion (Definition 2.29), this gap must be closed by a proof or by a precise reference to a relative extension theorem for piecewise smooth sections of open relations.","section":"Section 7 (Lemma 7.3)"},{"comment":"The proof begins with the assertion: 'Firstly, we approximate F by a section F′ that is piecewise smooth such that F′ is a solution on Op(M × ∂Δ^n) and such that F′ agrees with F on M × ∂Δ^n.' No argument is supplied for this step. This is load-bearing: Theorem 6.3 can only be applied to a piecewise smooth F′ that is already a solution on a neighborhood of the relative subcomplex Q = M × ∂Δ^n. The requirement F′ = F on the boundary is an exact interpolation condition, while the requirement that first-order jets in the collar lie in R is a differential condition; openness of R alone does not imply that an arbitrary piecewise-smooth approximation of F satisfies the latter. The same type of unproved extension appears in Lemma 7.3. A dedicated relative approximation/extension lemma for piecewise smooth sections over products M × Δ^n, with corners, is needed here.","section":"Section 7 (proof of Theorem 7.4)"}],"minor_comments":[{"comment":"In the paragraph handling simplices of a previous color, the inequality dC0(j1s(i), j1s(i+1)) < δ_i/2 does not follow from the stated item (1), which gives a δ_i/4 bound for an ε_{i+1}-slope perturbation on adjacent simplices. The indices and constants should be harmonized (probably δ_i/4), although this is a local fix and does not affect the overall strategy.","section":"Section 6.1 (proof of Theorem 6.1)"},{"comment":"The phrasing 'a triangulation K′ subdividing K1, K2, K' should be 'subdividing K1, K2, and K'. Also, the fact that the C^r distance depends on the auxiliary triangulation K is only explained after the definition; putting that caveat before the displayed formula would avoid confusion.","section":"Section 2.2.5 (Definition 2.12)"},{"comment":"The informal discussion of the topological-space analogue is useful motivation, but it contains the phrase 'we leave for the reader to explore'; if this is not intended as a mathematical claim, it is fine as a remark, but it should be clearly marked as non-essential and not as a lemma.","section":"Section 7.1"}],"recommendation":"major_revision","confidential_remarks":"The core jiggling theorem appears sound and is proved with commendable care. The missing relative extension/approximation step in Section 7 is a genuine gap, but it is local and likely repairable with standard techniques, so I recommend major revision rather than rejection. I have no concerns about novelty or attribution; the paper is a clear advance over Thurston's original jiggling lemma and should be encouraged."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the right generalization of Thurston's jiggling lemma, and the main jiggling theorem is in good shape. The advertised h-principle (Theorem 7.4) has a genuine gap in the proof as written, but it looks fillable rather than fatal.\n\nWhat's new: instead of jiggling a triangulation into general position against a distribution, they jiggle a section of any fiber bundle into a piecewise smooth solution of any open, fiberwise dense first-order relation. The packaging as a weak equivalence of simplicial sets is new, and the proof strategy really is different from Thurston's: the coloring argument avoids the Grassmannian estimates. The quantitative lemmas (3.8, 4.2, 4.6, 5.8) are worked out carefully, and the non-compact case gets real treatment rather than a shrug. The applications to immersions, contact forms and very general position are concrete and legitimate.\n\nSoft spots. The stress-test note is right: Theorem 7.4 depends on an unproved step. After a continuous section F on M × Δ^n with piecewise smooth solutions on the boundary, the proof says \"firstly, we approximate F by a section F' that is piecewise smooth...\" That approximation has to preserve the boundary values exactly and solve the relation on a collar, and for n > 1 the collar has corners. A standard relative smooth approximation argument plus openness of R should produce it, but it is not written down, and Lemma 7.3 has the same extension issue for horns. The paper even defines sSol_PS so that the Kan property reduces to this; it is the crux of the h-principle claim.\n\nThe manifold reduction in Theorem 6.3 is also condensed, especially the non-compact exhaustion argument. I don't think it's wrong, but a referee should ask for details.\n\nThe fiberwise density assumption is honestly discussed, including the conjectured rigidity for positive contact structures in dimension 3. That is a real limitation, not a hidden one.\n\nBottom line: this deserves a serious referee. The main theorem is valuable and the h-principle claim is likely true, but the proof of Theorem 7.4 needs a real fix. I'd send it out.","headline":"A genuinely new generalization of Thurston's jiggling with a solid main theorem, but the advertised simplicial-set h-principle has a gap in the write-up that needs patching.","tokens_in":34488,"tokens_out":2332,"would_cite":true,"duration_ms":21425,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57R45","57Q65","57R05","57R35"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every continuous section can be jiggled into a piecewise smooth solution of any open and fiberwise dense first-order differential relation.","keywords":["h-principle","jiggling","piecewise smooth sections","differential relations","crystalline subdivision","weak homotopy equivalence","contact forms","general position"],"falsifier":"Consider the immersion relation on maps $M \\to N$ with $\\dim M < \\dim N$, which is open and fiberwise dense, and a loop in $C^0(M,N)$ whose values are immersions; Theorem 7.4 predicts a piecewise immersive filling of that loop. A loop for which one can prove that no piecewise smooth immersion filling exists would falsify the central weak equivalence.","tokens_in":33376,"feed_emoji":"🧩","tokens_out":10102,"duration_ms":83103,"temperature":0.7,"pith_summary":"The paper claims a broad generalization of the classical jiggling lemma: for any open and fiberwise dense first-order differential relation $R \\subset J^1(E)$, every piecewise smooth section of the bundle $E$ can be subdivided and then $C^1$-perturbed into a piecewise smooth solution of $R$, while staying fixed on any region where it already solves $R$. The construction is parametric and relative, and it yields an h-principle without homotopical assumptions: the simplicial set of piecewise smooth solutions is weakly equivalent to the singular complex of all continuous sections. This matters because one mechanism then covers many geometric flexibility statements at once — transversality, immersions, submersions, piecewise contact forms, and general-position triangulations — on closed manifolds as well as open ones, where classical smooth h-principles often require formal data or fail.","feed_headline":"Jiggling turns any section into a piecewise solution","feed_subtitle":"For open, fiberwise dense first-order relations, piecewise solutions are as flexible as continuous sections.","key_machinery":"The engine is the jiggling construction of Section 6, which combines three ingredients. Crystalline subdivision refines a triangulation by pulling back the standard subdivision of a cube; its defining feature is that every simplex in every iterated subdivision is a scaled, translated copy of one of finitely many model simplices, so the quantities controlling $C^1$ estimates (edge length and the degeneracy constant $\\Lambda$) scale predictably with subdivision depth. A fixed finite coloring of the top-dimensional simplices allows the proof to perturb the section one color at a time, because same-colored simplices have disjoint stars and earlier solutions are only mildly disturbed. On each simplex, slope perturbation uses fiberwise density to tilt the jet of the section into the relation and then extends the tilted jet linearly, while a join construction glues the local tilts to the part of the section that is left unchanged. The parametric statement is encoded in the simplicial set $sSol_{PS}(R)$, whose $n$-simplices are piecewise solutions of the pulled-back relation on $M \\times \\Delta^n$.","core_discovery":"The central discovery is that jiggling is not just a way to put triangulations in general position: it is a way to solve differential relations. Given an open and fiberwise dense relation $R \\subset J^1(E)$, Theorem 6.3 takes any piecewise smooth section $s$ and any positive error function $\\varepsilon$ and produces an $\\varepsilon$-jiggling $(s', T')$ in which $s'$ is a piecewise smooth solution of $R$ and $s'$ agrees with $s$ on any prescribed subcomplex $Q$ where $s$ already solves $R$. Because the proof is relative and uniform over parameters, Theorem 7.4 follows: the inclusion of the simplicial set $sSol_{PS}(R)$ of piecewise solutions into $\\operatorname{Sing}(\\Gamma^0(E))$ is a weak homotopy equivalence of Kan complexes, meaning simplicial sets in which every horn can be filled. The authors read this as an h-principle without homotopical assumptions, since the first jet of a piecewise solution may jump along the triangulation and hence no formal solution or jet homotopy class appears in the statement.","pith_inferences":["Editorial inference: if Theorem 7.4 is correct, then in the piecewise smooth category every open and fiberwise dense first-order relation is automatically flexible at the level of homotopy types; rigidity phenomena, such as the conjectured rigidity of positive contact structures in dimension 3, would have to come precisely from failure of fiberwise density.","Editorial inference: the color-by-color induction is the transferable core of the proof, since it replaces delicate Grassmannian distance estimates with a finite coloring; a testable extension is to higher-order jets, where a subdivision scheme controlling higher derivatives might yield analogous h-principles without homotopical assumptions for relations in $J^r(E)$.","Editorial inference: one could try to replace fiberwise density by an ampleness condition in the sense of convex integration; if sufficiently controlled local tilts exist without density, jiggling would extend piecewise h-principles to a larger class of relations, including some that are not fiberwise dense."],"forward_implications":["Every continuous section of $E$, and every continuous family of sections parametrized by a simplex, can be deformed relatively into a piecewise smooth solution of $R$; in simplicial terms, $sSol_{PS}(R)$ and $\\operatorname{Sing}(\\Gamma^0(E))$ have the same weak homotopy type.","Any map between manifolds can be jiggled to be piecewise transverse to a fixed distribution, and, in the appropriate dimensions, to be piecewise immersive or submersive; the corresponding inclusions of solution spaces into spaces of continuous maps are weak homotopy equivalences.","On any odd-dimensional manifold, every nonvanishing 1-form is $C^1$-close to a piecewise smooth contact form, and the simplicial set of piecewise contact forms is weakly equivalent to the space of nonvanishing continuous 1-forms, on closed manifolds as well as open ones.","Every triangulation can be jiggled to be in very general position with respect to a distribution, including on noncompact manifolds, and the same statement holds for finitely many distributions at once."],"supporting_citations":[{"why":"Supplies the original jiggling lemma and the crystalline subdivision scheme that the proof generalizes.","marker":"[31]"},{"why":"Provides the coloring argument that lets the proof perturb many simplices at once without interference.","marker":"[9]"},{"why":"Gives the extension of triangulations of $M \\times \\partial \\Delta^n$ to $M \\times \\Delta^n$ used in the parametric h-principle.","marker":"[21]"},{"why":"Sets up the simplicial sets and Kan-complex machinery in which Theorem 7.4 is stated and proved.","marker":"[16]"},{"why":"Provides the classical h-principle framework that the paper's no-homotopy-assumptions version is compared with.","marker":"[15]"}],"fun_headline_variants":["Jiggling converts any section into a solution","No homotopy needed: jiggling yields solutions","Jiggling: solutions from any section","Any section can be jiggled into a solution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is fiberwise density: at every point of the bundle, the allowed first-order jets must be dense among all possible slopes over that point, so that any jet can be tilted into the relation by an arbitrarily small change of slope.","fun_headline_variants_meta":{"raw":{"variants":["Jiggling converts any section into a solution","No homotopy needed: jiggling yields solutions","Jiggling: solutions from any section","Any section can be jiggled into a solution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0003,"raw_usage":{"total_tokens":1704,"prompt_tokens":888,"completion_tokens":816,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":504,"completion_tokens_details":{"reasoning_tokens":755}},"tokens_in":504,"tokens_out":816,"duration_ms":7158,"temperature":1.0,"reasoning_tokens":755,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T15:46:12.396278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Consider the immersion relation on maps $M \\to N$ with $\\dim M < \\dim N$, which is open and fiberwise dense, and a loop in $C^0(M,N)$ whose values are immersions; Theorem 7.4 predicts a piecewise immersive filling of that loop. A loop for which one can prove that no piecewise smooth immersion filling exists would falsify the central weak equivalence.","supporting_citations":[{"cited_title":"Symplectic submanifolds and almost-complex geometry","cited_arxiv_id":null,"evidence_quote":"Provides the coloring argument that lets the proof perturb many simplices at once without interference."},{"cited_title":"Heuts and I","cited_arxiv_id":null,"evidence_quote":"Sets up the simplicial sets and Kan-complex machinery in which Theorem 7.4 is stated and proved."},{"cited_title":"Stable mappings of foliations into manifolds","cited_arxiv_id":null,"evidence_quote":"Provides the classical h-principle framework that the paper's no-homotopy-assumptions version is compared with."}],"review_version":1}