{"id":"a149e4cd-d4de-45eb-b9c4-d2b1559f25e6","arxiv_id":"2412.12713","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A Sobolev trace on a collar extends if and only if its restriction to every member of some finite open covering extends, with a controlled energy bound.","lead":"The paper proves that a boundary map between two compact curved spaces can be filled in by a finite-energy Sobolev map whenever it can be filled in on each piece of a finite open covering. This local-to-global principle simplifies the study of trace extensions for manifold-valued Sobolev maps and yields new linear energy estimates.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The local Theorem 1.1 is convincingly supported by the folding argument, but the global Theorem 1.4 depends on an unpublished companion [9] for essential steps, leaving the advertised global consequence unverified.","rationale":"The reader's verdict is CONDITIONAL, and the reader's weakest assumption correctly identifies the dependence of Theorem 1.4 on the unpublished companion [9] as the main unverified support. My independent reading of Proposition 2.1 and Proposition 2.2 confirms that the trace agreement is preserved by the folding construction: the internal interface conditions are exactly the trace equality (2.2), and the gluing theorem applies to N-valued Sobolev maps because they are embedded in R^ν. The chart reduction for compact manifolds and arbitrary finite open covers is a standard refinement argument; while it is not written out, it does not threaten the local theorem. No internal inconsistency appears in the local proof. The single load-bearing concern is therefore the external, unpublished dependency in the proof of Theorem 1.4. This does not lower confidence in Theorem 1.1, but it does justify the reader's CONDITIONAL verdict until [9] is available and the cited propositions are verified, or the global proof is made self-contained.","tokens_in":13704,"tokens_out":22714,"duration_ms":197402,"concrete_test":"Obtain the companion manuscript [9] and verify that Prop. 9.4, Prop. 9.11, and Cor. 9.10 are stated with matching hypotheses and complete proofs; specifically check that Cor. 9.10 supplies the homotopy V∘ξ|Σ0 for the piecewise-affine map ξ under the VMO hypotheses used in the step (iii)→(i). If [9] is not available, independently re-derive the (iii)→(i) implication of Theorem 1.4 using only [27] and [35]; if the construction of Φ and ξ cannot be completed, the global claim should be marked conditional.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central local claim (Theorem 1.1) is established by Proposition 2.2. The trace-agreement condition (2.2) is indeed preserved by the piecewise linear changes of variables: on Γ0 both sides pull back to tr U0 = tr U1 on W×(0,1)×{0}, and on Γ1 both sides are the same trace of U1; the gluing step uses the standard Sobolev gluing theorem [26, Th. 18.1] and is valid for manifold-valued maps. The chart/refinement reduction for arbitrary open coverings is standard, though it should be stated explicitly. The load-bearing gap is in the global part: Theorem 1.4's proof invokes the unpublished companion [9, Prop. 9.4, Prop. 9.11, Cor. 9.10] for the topological screening arguments that convert a global homotopy condition into the Sobolev-extension criterion and that identify the homotopy class of u∘Φ∘ξ|Σ0 with V∘ξ|Σ0. Reference [9] is listed as 'in preparation', so these results cannot currently be checked. Without them, the equivalences (i)⇔(ii)⇔(iii) and the energy estimate (1.12) in Theorem 1.4 are not supported. The abstract's final statement that 'analytical obstructions ... are purely local' therefore rests on an unverified external dependency, not on an error in the local proof.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper establishes a locality principle for the trace extension problem for Sobolev maps between compact Riemannian manifolds. Theorem 1.1 states that a Borel map u:M' to N is the trace on M' of a W^{1,p}-map on the collar M' times [0,1) if and only if its restriction to each member of a finite open cover extends locally, and the global extension energy is controlled by the sum of the local extension energies. The proof uses an explicit folding of the domain (Proposition 2.1), an inductive gluing over charts (Proposition 2.2), and a cone-covering lemma (Lemma 2.3). Corollaries include a linear boundedness principle (Theorem 1.2) and an extension for penalized energies (Theorem 1.3). In the global boundary case, Theorem 1.4 characterizes traceability from the boundary of a manifold by collar traceability plus homotopy conditions and gives a convex energy estimate, following the author's earlier screening characterization [27] and an unpublished companion [9].","tokens_in":14048,"tokens_out":9600,"duration_ms":79798,"significance":"The local theorem is an elegant reduction: it shows that trace extension and its energy scale are local in the domain, so future characterizations can be restricted to balls. The folding construction is explicit and avoids linear partition-of-unity arguments that fail for manifold-valued maps, and the quantitative estimate with constant depending only on the cover and p is valuable. The main caveat is that the global Theorem 1.4, and hence the abstract's claim that analytical obstructions are purely local, rests on statements in an 'in preparation' companion paper [9] that cannot currently be checked. The local part is nonetheless a self-contained contribution.","major_comments":[{"comment":"The equivalence (i) iff (ii) iff (iii) and the energy estimate (1.12) are proved by invoking [9, Prop. 9.4, proof of Pr. 9.11, Cor. 9.10], and [9] is listed as 'in preparation' with no preprint or detailed statements available. These invocations are load-bearing: without them, the global characterization is not established. Please either provide complete proofs of the quoted results, state Theorem 1.4 as conditional on [9], or restrict the abstract's global conclusion to the parts that are actually proved in this manuscript.","section":"Section 3, proof of Theorem 1.4"}],"minor_comments":[{"comment":"The statement of Proposition 2.2 contains a typo: 'there exists U in dot W^{1,p}(G_i times (0,1), N)' should read 'there exists U in dot W^{1,p}(M' times (0,1), N)', and in the compatibility condition 'u_j' should be 'U_j'.","section":"Section 2.2, Proposition 2.2"},{"comment":"In the sentence listing the three pieces, the second membership for Sigma-sharp is correct, but the third should be 'U*|_{Sigma_1} in dot W^{1,p}(W times Sigma_1, N)' rather than 'U*|_{Sigma_sharp} in dot W^{1,p}(W times Sigma_1, N)'.","section":"Section 2.1, proof of Proposition 2.1"},{"comment":"The displayed identity for (C_i cap closed ball) union (E_i setminus closed ball) contains an undefined set 'G_i' in R^m, and the claimed openness of B_{r_i} union (C_i cap closed ball) union (E_i setminus closed ball) is not immediate from the displayed algebra; please rewrite this step with precise set operations and a clear verification of openness.","section":"Section 2.2, proof of Proposition 2.2"},{"comment":"Reference [18] appears in the bibliography but is not cited in the text; either cite it where relevant or remove it.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The local part of the paper (Theorems 1.1-1.3) is, in my reading, correct and publishable. The global part is currently conditional because of the dependence on the unpublished companion [9]. I would advise making acceptance contingent on either making the companion available or reframing the paper around the local theorem. The overlap with the author's own prior work [27] and [35] is significant, but the local-to-global transfer via folding is new and appears sound."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is Theorem 1.1: trace extension and the extension energy for manifold-valued Sobolev maps localize to any finite open cover. That was known only for linear targets, and the paper gives a nonlinear proof by folding, which is exactly the right tool. The energy estimate with constant depending only on the cover is what you'd want, and Theorem 1.3 extends it to penalized energies at no extra cost.\n\nProposition 2.1 is clean. The gluing condition is explicit, and the trace agreement on the fold lines is really preserved: on Gamma0 both sides pull back to the common trace, on Gamma1 both sides are the trace of U1. The stress-test worry that the piecewise linear changes of variable break the trace agreement does not land. Proposition 2.2 is more compressed: the induction over the cover using Lemma 2.3 and the cone is sketched, and there are typographical slips in the set displays, but the construction is standard and the ingredients are visible.\n\nThe real soft spot is Theorem 1.4. The proof delegates to the unpublished companion [9] for Prop. 9.4, Prop. 9.11, and Cor. 9.10. These are load-bearing for converting global homotopy conditions into the Sobolev extension criterion and for the homotopy identification. Since [9] is listed as 'in preparation', the global equivalences and the energy estimate (1.12) cannot currently be checked. This does not touch Theorem 1.1, whose proof is self-contained, but the abstract's last sentence about 'purely local' analytical obstructions is only supported conditionally on [9].\n\nMinor issues: 'Riemnnian' typo in the introduction, and some notational errors in the proof of Proposition 2.2. These are minor.\n\nWho is this for: people in trace theory for Sobolev maps between manifolds, and anyone using extension theorems in calculus of variations. The local theorem is a genuine advance and deserves refereeing. The referee should ask for a self-contained proof of Theorem 1.4, or at least an explicit statement that the global part depends on [9], and for a cleaned-up Proposition 2.2.\n\nRecommendation: send to peer review. It is a serious paper with a real theorem; conditional acceptance on making the global part verifiable is appropriate.","headline":"Solid new local trace-extension theorem with a clean folding proof; the global theorem is unverifiable until the unpublished companion appears.","tokens_in":14521,"tokens_out":1750,"would_cite":true,"duration_ms":14999,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["58D15","46E35","46T10","58C25","58J32"],"pacs":[],"model":"deepseek-v4-flash","headline":"A Sobolev trace extension is local: finite coverings decide","keywords":["Extension of traces","Sobolev mappings","Trace theory","Sobolev-Slobodecki spaces","Folding construction","Local character","Collar neighbourhood"],"falsifier":"Take a flat model with $W$ a point, so the domain is the unit square, and pick two $\\dot W^{1,p}$ extensions of the same trace on the bottom edge that are smooth except for a concentration along an internal curve; compute $U_*$ by formula (2.1) and test whether its energy is bounded by the sum of the two energies and whether its trace on the bottom edge equals the common trace. A violation for some exponent $p$ (for instance $p=1$) or for some trace with a jump across the overlap would refute Proposition 2.1, and hence the local character theorem.","tokens_in":13497,"feed_emoji":"🧩","tokens_out":10779,"duration_ms":86884,"temperature":0.7,"pith_summary":"The paper proves that the Sobolev trace extension property on a collar neighborhood is local in the domain: a map from a compact manifold into a target manifold is the trace of a Sobolev map from the manifold times an interval exactly when each restriction to a member of some finite open covering is such a trace, and the minimal extension energy on the whole domain is controlled by the sum of the local energies. This matters because it lets analysts check extendability patch by patch on small coordinate balls, where the target manifold's topology enters only through the local model. The proof avoids linear averaging, which does not respect manifold-valued constraints, and instead uses a folding construction that glues two local extensions with matching traces. A corollary gives a general linear energy estimate whenever the trace is surjective, and a global version separates the collar extension from a purely topological obstruction on a skeleton of the boundary.","feed_headline":"Sobolev traces extend globally if they extend on each patch","feed_subtitle":"A boundary datum extends into a collar iff each term of a finite open cover does, with comparable minimal energy.","key_machinery":"The load-bearing construction is the folding map $U_*$ defined in Proposition 2.1. Given two Sobolev extensions $U_0$ and $U_1$ on $W\\times(0,1)\\times(0,1)$ with equal traces on $W\\times(0,1)\\times\\{0\\}$, $U_*$ is set to $U_0$ on the region below a diagonal (after the substitution $(x',2x_{m-1},x_m-2x_{m-1})$), to $U_1$ on the wedge between the diagonal and the horizontal bisector (after substitution $(x',x_m,2x_{m-1}-x_m)$), and to $U_1$ unchanged above the diagonal. The two pieces agree on the interfaces precisely because the original traces agree, so the gluing property of Sobolev functions gives $U_*\\in\\dot W^{1,p}(W\\times(0,1)\\times(0,1),N)$ with energy bounded by the sum of the two energies. Proposition 2.2 turns this local folding into a global gluing over a finite covering through an induction that uses a cone-type separation lemma (Lemma 2.3) to fit a new open set between a closed remaining region and the previously glued region.","core_discovery":"The central claim is Theorem 1.1: for compact Riemannian manifolds $M'$ and $N$, a Borel-measurable map $u\\colon M'\\to N$ belongs to $\\operatorname{tr}_{M'}(\\dot W^{1,p}(M'\\times[0,1),N))$ if and only if for each $i$ in a finite open covering $(G_i)$ of $M'$, the restriction $u|_{G_i}$ belongs to $\\operatorname{tr}_{G_i}(\\dot W^{1,p}(G_i\\times[0,1),N))$, and the extension energy satisfies $E^1_{\\mathrm{ext}}(u,M',M'\\times[0,1))\\le C\\sum_i E^1_{\\mathrm{ext}}(u|_{G_i},G_i,G_i\\times[0,1))$, where $C$ depends only on the covering and $p$. The paper also proves the same locality for penalized extension energies (Theorem 1.3) and for the global extension problem from the boundary of a manifold (Theorem 1.4), where the only additional condition is a homotopy-theoretic test for the boundary datum on a skeleton of the manifold. In effect, the analytical obstructions to extending a boundary datum into a collar are purely local; global topology of the domain does not enter the collar extension.","pith_inferences":["The folding construction resembles a nonlinear replacement for convex combination that stays inside the target manifold; it may transfer to other constrained settings where averaging violates the constraint, such as maps into stratified spaces or with symmetry constraints.","The locality result makes traceability a checkable finite procedure on a fixed atlas: once a finite covering and charts are chosen, the energy constant in Theorem 1.1 is an explicit quantity of the covering, so numerical certificates of local extendability would certify global extendability.","Quantifying how the constant in Theorem 1.1 depends on the covering (through the chart bilipschitz constants and the radii from Lemma 2.3) would make the estimate ready for computational use; the paper only proves existence of such a constant.","The endpoint case $p=1$, where the trace space is $L^1$-based, may need a separate check: the folding estimates hold formally, but the gluing theory for traces at $p=1$ has different features, so the comparability of energies could behave differently."],"forward_implications":["To determine whether a map on a compact manifold is a Sobolev trace on a collar, it is enough to test each set of some finite open covering; the energy bound is uniform once the covering is fixed.","Wherever the trace is surjective, the minimal extension energy is bounded by a constant times the Gagliardo energy, so surjectivity automatically comes with the sharp linear estimate (Theorem 1.2).","The global extension problem for a manifold with boundary reduces to a local collar extension plus a topological condition on the boundary datum over a skeleton of the manifold (Theorem 1.4).","Penalized extension energies used in relaxation approaches satisfy the same local energy bound (Theorem 1.3), so localization can be applied to variational approximations."],"supporting_citations":[{"why":"Prior characterization of surjective traces on collars and of global extension, which Theorem 1.1 refines and Theorem 1.4 builds upon.","marker":"[35]"},{"why":"Quantitative topological screening characterization of traces, used in the proof of Theorem 1.4.","marker":"[27]"},{"why":"Unpublished companion work supplying Propositions 9.4, 9.11 and Corollary 9.10 used in the proof of Theorem 1.4.","marker":"[9]"},{"why":"Provides the gluing property of Sobolev functions with the same trace, which underlies the folding construction in Proposition 2.1.","marker":"[26]"},{"why":"Uniform boundedness principle for extension of traces, combined with Theorem 1.1 to prove Theorem 1.2.","marker":"[31]"}],"fun_headline_variants":["Sobolev trace extension is local: cover patches decide","If a Sobolev trace extends locally, it extends globally","Trace extension iff each open cover patch extends","Boundary traces extend iff each local patch does"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that two local extensions that agree as traces on an overlap can be folded, after piecewise-linear changes of variables and passage through manifold charts, into one Sobolev extension whose trace equals both, and that the global theorem's unpublished companion supplies three technical propositions as stated; if either fails, the gluing induction collapses.","fun_headline_variants_meta":{"raw":{"variants":["Sobolev trace extension is local: cover patches decide","If a Sobolev trace extends locally, it extends globally","Trace extension iff each open cover patch extends","Boundary traces extend iff each local patch does"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000251,"raw_usage":{"total_tokens":1561,"prompt_tokens":952,"completion_tokens":609,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":568,"completion_tokens_details":{"reasoning_tokens":545}},"tokens_in":568,"tokens_out":609,"duration_ms":5507,"temperature":1.0,"reasoning_tokens":545,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:49:49.674299+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a flat model with $W$ a point, so the domain is the unit square, and pick two $\\dot W^{1,p}$ extensions of the same trace on the bottom edge that are smooth except for a concentration along an internal curve; compute $U_*$ by formula (2.1) and test whether its energy is bounded by the sum of the two energies and whether its trace on the bottom edge equals the common trace. A violation for some exponent $p$ (for instance $p=1$) or for some trace with a jump across the overlap would refute Proposition 2.1, and hence the local character theorem.","supporting_citations":[{"cited_title":"Mazowiecka and J","cited_arxiv_id":null,"evidence_quote":"Quantitative topological screening characterization of traces, used in the proof of Theorem 1.4."},{"cited_title":"Bousquet, A","cited_arxiv_id":null,"evidence_quote":"Unpublished companion work supplying Propositions 9.4, 9.11 and Corollary 9.10 used in the proof of Theorem 1.4."}],"review_version":1}