{"id":"3012f2b0-1bfc-4176-a75b-78c730bbe935","arxiv_id":"2512.12305","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Continuous functions on the unit disk with the strong maximum principle and a zero at the center have nodal sets of length at least 2, and the constant 2 is sharp (attained by u = x1).","lead":"This mathematics paper proves that any continuous function of two variables that obeys a strong maximum principle and vanishes at the origin must have a zero set inside the unit disk that is at least as long as the disk's diameter, with the bound attained by the simplest example. It also supplies new lower bounds for zero sets of fast-oscillating elliptic equations with periodic coefficients, a setting where only upper bounds were known.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.9's proof hinges on an unjustified rotation step in Claim 2; the claim can be repaired by a 1-Lipschitz radial projection, so the central result is likely correct but needs revision.","rationale":"The reader's conditional verdict is appropriate. The most load-bearing concern is the rotation-covering step in Claim 2, which is indeed unjustified as written and is essential to the proof of Theorem 1.9. However, the fix is straightforward: replace the rotation by the radial projection F(r,θ)=(r,θ_mid), which is 1-Lipschitz. The two-zeros-per-circle premise the reader identified as fragile is actually a consequence of the strong maximum principle, so I do not fully agree with the reader's framing. The rest of the proof of Theorem 1.9—the closedness of I and the left-neighborhood extension—is sound. The homogenization theorems (1.5–1.7) have additional issues as the reader notes, but those are secondary to the paper's central claim. Since the main theorem can be repaired without changing its content, and the paper as written has a genuine gap, a conditional verdict is correct; my review does not alter the reader's conclusion.","tokens_in":12141,"tokens_out":18259,"duration_ms":156010,"concrete_test":"Rewrite the covering step (4.13)-(4.21) replacing the rotation with the radial projection F(r,θ)=(r,θ_mid). Check that for any cover of Z1 by balls B_i^1 with centers x_i^1, the projected balls B(F(x_i^1), r_i^1) cover l1 and have the same total diameter sum; use the 1-Lipschitz property to conclude H^1(Z1) ≥ H^1(l1). If this derivation succeeds, Claim 2 holds and Theorem 1.9 is proved. If it fails, the theorem is unproved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 4, Claim 2 (equations (4.13)-(4.21)), the proof aims to show that any covering of Z1 by balls B_i^1 of radii r_i^1 ≤ r_s/10 yields a covering of the radial segment l1 by the rotated balls \\tilde B_i^1 centered at (|x_i^1|, (θ1+θ2)/2). The inference \"if z_t ∈ B_i^1, then (t, (θ1+θ2)/2) ∈ \\tilde B_i^1\" is not justified: rotating only the center along its circle is not an isometry of the ball containing z_t unless the point is also rotated by the same angle. For a point z_t=(t,φ) with φ far from the mid-angle, the distance between z_t and its mid-angle image can be as large as 2, while r_i^1 ≤ r_s/10 is small; the rotated ball need not contain the mid-angle point. Thus the covering argument, as written, is invalid. However, the difficulty is local and repairable: the map F(r,θ)=(r,θ_mid) is 1-Lipschitz, maps Z1 into l1, and every circle in the annulus contains a zero in the relevant sector, so F(Z1) contains the full segment l1. Hence H^1(l1) ≤ H^1(F(Z1)) ≤ H^1(Z1), and similarly H^1(l2) ≤ H^1(Z2). This restores the desired lower bound H^1 ≥ r_s. The related premise that each circle ∂B_s contains at least two zeros is actually forced by the strong maximum principle and is not the fragile point.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies lower bounds for nodal sets Z(u)={u=0}. Theorem 1.5 states that for divergence-form periodic homogenization in n≥3, a nontrivial solution u_ε with u_ε(0)=0 and doubling index N_0 satisfies H^{n-1}(Z(u_ε)∩B_1)≥C/N_0^{n-1} for ε<ε_0(A,n,N_0); Theorem 1.6 upgrades this to a uniform C(N_0) bound. Theorem 1.7 gives a constant lower bound in n=2 independent of ε and N_0, and Theorem 3.2 gives a similar constant bound for elliptic equations with C^1 coefficients, independent of the Lipschitz constant. The central result, Theorem 1.9, asserts that every continuous function on B_1⊂R^2 satisfying the strong maximum principle with u(0)=0 has H^1(Z(u)∩B_1)≥2, with the bound sharp for u=x_1; Theorem 4.3 extends this to the weak SMP class. The proof uses the set I of radii s for which the nodal set in the annulus T_{s,1} has length at least 2(1-s), and proves I=[0,1] via closure and a left-propagation claim.","tokens_in":12384,"tokens_out":21833,"duration_ms":184682,"significance":"If the gaps are repaired, the paper would establish a genuinely new phenomenon in two dimensions: the nodal lower bound is a consequence of the maximum principle rather than of PDE structure, with an optimal constant 2. The homogenization results respond to a natural question, and the dependence C(N_0) in Theorem 1.6 is in line with the known almost-sharp bounds from LLPS24 and KZZ22. The proof is analytic and self-contained apart from standard cited theorems; it contains no fitted parameters, and the sharp example is explicit. However, as written, the proof of Theorem 1.9 has a gap in the covering step (Claim 2), and several supporting arguments in Sections 2-3 require additional justifications. The overall significance is high if the revision succeeds.","major_comments":[{"comment":"The covering argument in Claim 2 is invalid as written. From z_t=(t,φ)∈B_i^1 it does not follow that (t,(θ1+θ2)/2) lies in the rotated ball \\tilde B_i^1 centered at (|x_i^1|,(θ1+θ2)/2); only the center was rotated, not z_t, and for φ near a boundary ray the distance between z_t and the mid-angle point can be O(1), far larger than r_i^1≤r_s/10. Thus (4.18) and the inclusion (4.19) are unjustified, and the subsequent 'by the definition of Hausdorff measure' step does not supply the missing comparison. A repair is available: the radial projection F(r,θ)=(r,(θ1+θ2)/2) is 1-Lipschitz, and by (4.16) F maps the zero set in the sector onto the segment l1, so H^1(l1)≤H^1(Z1); similarly for l2. This gives the desired H^1≥r_s. Because this is the central step of Theorem 1.9, the proof must be rewritten.","section":"Section 4, Claim 2 (Eqs. (4.13)-(4.21))"},{"comment":"In the decreasing-subsequence case of Claim 1, the printed chain H^1(T_{s,1})≥H^1(T_{s_nk,1})≥2(1-s_nk) is correct but insufficient: since 2(1-s_nk)≤2(1-s), it does not imply H^1(T_{s,1})≥2(1-s). The conclusion follows from the stronger fact that T_{s_nk,1}↑T_{s,1} and H^1(T_{s,1})=lim_k H^1(T_{s_nk,1}) (using finiteness of the measure). The claim itself is true, but the proof has a logical gap at (4.6)-(4.7).","section":"Section 4, Claim 1 (Eqs. (4.5)-(4.7))"},{"comment":"In Lemma 2.4, after obtaining at least k/3 indices that are both S+-good and S--good, the proof adds H^{n-1}≥C/N^{n-1} over these indices without ensuring that the corresponding balls/annuli are disjoint; adjacent annuli B_{r_i+1}\\setminus B_{r_i-1} overlap. A selection of every third index makes them disjoint. In addition, the 'standard nodal sets estimate' used at (2.23) for an arbitrary continuous u needs a precise statement or reference. Since this lemma is the engine for Theorems 1.5-1.6, the section should be made rigorous.","section":"Section 2, Lemma 2.4 (Eqs. (2.18)-(2.23))"},{"comment":"In the proof of Theorem 3.2, the assertion 'by Lemma 3.7' at (3.46) is not enough: Lemma 3.7 is a doubling-ratio inequality, not a measure lower bound. The H^1 bound in B(z_i,1/2) follows from Theorem 3.8 after scaling the equation to unit ball, and the disjointness of the balls (distance at least |i-j|≥1) should be stated before summing over i. Also, the passage 'Combining Theorem 1.2' to obtain Theorem 1.6 for ε≥ε0 is not elaborated; it likely requires the C^1-coefficient version of Theorem 1.2 and a bound on the Lipschitz constant by 1/ε0. Please make these steps explicit.","section":"Section 3, Theorem 3.2 proof and Theorem 1.6"}],"minor_comments":[{"comment":"Typos include 'duobling', 'exsits', 'osciallating', 'conseqence', 'unknow', and 'denpending'. Please proofread.","section":"Throughout"},{"comment":"At (3.45), 'mean value theorem' should be 'intermediate value theorem' when concluding the existence of z_i∈∂B_i with v(z_i)=0.","section":"Section 3, Theorem 3.2 proof"},{"comment":"Specify that r_s is chosen small enough (e.g. r_s<min(1-s, ...)) so the sectors and balls used in the annulus lie in B_1; the proof implicitly needs this.","section":"Section 4, Claim 2"},{"comment":"The chain H^{n-1}≥C ε^{n-1}≥C hides the fact that the second C depends on the fixed lower bound for ε; rewrite with explicit constants.","section":"Section 3, Theorem 1.7 Case 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is a promising short note. The main theorem is likely correct, and the fix via a 1-Lipschitz projection is local. I recommend major revision rather than rejection. Please ask the authors to also clarify the relation to [L WY25] and to make the proof of Theorem 1.6 self-contained."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's real contribution is Theorem 1.9: any continuous function on the unit disk satisfying the strong maximum principle, with u(0)=0, has nodal length at least 2, and the constant is sharp. That is a genuinely new result — nothing in the cited PDE-specific literature gives this for arbitrary SMP functions, and it identifies a maximum-principle phenomenon rather than a PDE-structural one. The proof idea is also appealing: two crossings per concentric circle, converted into a length lower bound. The authors are honest that the homogenization bounds are far from optimal, and I see no circularity or fitted parameters; the heavy inputs come from LLPS24 and KZZ22.\n\nThe soft spots are real but mostly fixable. The rotation-covering step in Claim 2 is not justified as written: rotating the center of a ball along a circle is not an isometry of the ball, so the claimed containment can fail badly. The stress-test note is right that a 1-Lipschitz radial projection F(r,θ)=(r,θ_mid) repairs this, mapping each zero set into the corresponding radial segment and preserving the length comparison. That makes the central argument recoverable. Claim 1 also has a reversed inequality in the decreasing-subsequence case, but that is a minor fix via upper continuity of H^1.\n\nThe homogenization half is rougher. Theorem 1.5's counting gives the wrong exponent as printed; Theorem 1.6 asserts an epsilon-uniform C(N0) without covering the regime epsilon >= epsilon0(N0) where the KZZ22 approximation premise fails and LLPS24 constants degrade; Theorem 3.2 is stated for C^1 coefficients but the proof assumes Lipschitz constants and blows up; Theorem 3.9 is asserted without proof. These are load-bearing for the homogenization claims, and they would need to be addressed before those theorems can be used. The authors do flag that the quantitative form in Theorem 3.9 is open, which is good, but the other gaps are not flagged.\n\nOverall, the central SMP theorem is likely correct and deserves to be published after repair. The homogenization results are suggestive but need more work. This is a paper for nodal-set and homogenization people, and it deserves a serious referee rather than a desk reject. I would send it to peer review with the expectation of revision.","headline":"The sharp 2D strong-maximum-principle nodal lower bound is new and likely correct, but the proof as written has a real gap in the rotation argument and the homogenization half needs tightening.","tokens_in":13077,"tokens_out":1486,"would_cite":true,"duration_ms":14605,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35B27","35B50","35J15","28A75"],"pacs":[],"model":"deepseek-v4-flash","headline":"A continuous function on the unit disk that vanishes at the origin and satisfies the strong maximum principle must have a zero set of length at least 2.","keywords":["nodal sets","lower bounds","elliptic homogenization","strong maximum principle","Hausdorff measure","two-dimensional","doubling index","periodic coefficients"],"falsifier":"Construct a continuous function on the unit disk, vanishing at the origin and satisfying the strong maximum principle, whose zero set intersects each circle centered at the origin at most once (or whose total length is less than 2). For instance, a function with a single smooth zero curve spiraling from the origin to the boundary would, if it existed, refute Theorem 1.9. Alternatively, verifying the covering/rotation inequality (4.13)–(4.21) on a simple piecewise-linear example would test the geometric step.","tokens_in":11836,"feed_emoji":"📏","tokens_out":21899,"duration_ms":191162,"temperature":0.7,"pith_summary":"The paper establishes that in two dimensions, the presence of nodal volume is a pure maximum-principle phenomenon. It proves a uniform lower bound for nodal volume in elliptic homogenization (depending only on the doubling index in dimensions n≥3, and constant in dimension two), and then goes further: any continuous function on the unit disk that satisfies the strong maximum principle and vanishes at the origin has a zero set of length at least 2. The constant 2 is sharp, attained by a linear function. This transfers the question of lower nodal bounds from the analytic structure of PDEs to a topological constraint on sign changes.","feed_headline":"Nodal sets in 2D always have length at least 2","feed_subtitle":"Even without any equation—continuity plus the maximum principle forces a zero set of length at least 2.","key_machinery":"The key mechanism is the forced double sign change on every concentric circle: for an SMP function with u(0)=0, the ball B_s must contain both positive and negative values, so by continuity the boundary ∂B_s contains at least two zeros of u. The proof converts this into an H^1 lower bound on the annulus through a covering/rotation argument that maps the two zero arcs onto two fixed radial segments, yielding at least r_s of length per annulus, and then propagates the bound from s=1 down to s=0 by a closedness argument. In the homogenization part, the load-bearing tools are a harmonic approximation theorem (Theorem 2.2) that replaces the oscillating solution by a harmonic function when ε ≪ N^{","core_discovery":"The central discovery is Theorem 1.9: for every continuous u in B(0,1) ⊂ R^2 satisfying the strong maximum principle with u(0)=0, the 1-dimensional Hausdorff measure of the zero set in the unit ball is at least 2. The proof shows that on every circle centered at the origin, the strong maximum principle forces u to take both positive and negative values, so each circle carries at least two zeros; the argument then converts this 'two zeros per circle' fact into a lower bound on the Hausdorff measure of the full zero set. The companion results for elliptic homogenization (Theorems 1.5–1.7) establish lower bounds for oscillating-coefficient equations that are uniform in the oscillation scale, an","pith_inferences":["The two-zeros-per-circle property, together with the standard coarea inequality for Hausdorff measures, would prove Theorem 1.9 directly without the covering/rotation argument; the paper's elaborate geometric step may be unnecessary.","If Theorem 1.9 extends to higher dimensions, one would expect H^{n-1}({u=0}) ≥ c(n) > 0 for SMP functions vanishing at the origin, but the circle argument does not obviously generalize to spheres.","The homogenization lower bound C(N_0) degrades as N_0 grows; the conjectured fully uniform bound independent of N_0 would require a mechanism beyond harmonic approximation, which currently demands ε ≪ N^{-1/2}.","A natural test of the geometric step is to compute the nodal length of piecewise-linear SMP functions with polygonal zero sets; if any such function achieves length < 2, Theorem 1.9 would fail."],"forward_implications":["In dimension two, the nodal lower bound for elliptic homogenization is independent of both the oscillation scale ε and the doubling index; it depends only on the ellipticity constant (Theorem 1.7).","For any uniformly elliptic C^1 operator in two dimensions, a nontrivial solution vanishing at a point has nodal length at least a constant depending only on ellipticity (Theorem 3.2).","The sharp constant 2 in Theorem 1.9 shows that no additional PDE structure can improve the lower bound: a linear function already attains it.","In dimensions n≥3, homogenization solutions with a bounded doubling index N_0 have nodal volume bounded below by a constant depending on N_0, uniformly in ε (Theorem 1.6).","The results imply that any equation whose solutions obey the strong maximum principle inherits a constant nodal lower bound in 2D, regardless of coefficient regularity or oscillation."],"fun_headline_variants":["Strong max principle forces nodal length at least 2","Two zeros per circle: max principle implies length ≥2","No PDE needed: strong max principle yields nodal length ≥2","Homogenization: nodal length ≥2 uniformly in 2D","Max principle alone: zero set length ≥2 in 2D"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that each circle ∂B_s carries at least two zeros of u, together with the geometric conversion of that fact into a Hausdorff-measure lower bound, a step that the manuscript does not fully justify.","fun_headline_variants_meta":{"raw":{"variants":["Strong max principle forces nodal length at least 2","Two zeros per circle: max principle implies length ≥2","No PDE needed: strong max principle yields nodal length ≥2","Homogenization: nodal length ≥2 uniformly in 2D","Max principle alone: zero set length ≥2 in 2D"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000903,"raw_usage":{"total_tokens":3662,"prompt_tokens":623,"completion_tokens":3039,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":367,"completion_tokens_details":{"reasoning_tokens":2954}},"tokens_in":367,"tokens_out":3039,"duration_ms":19790,"temperature":1.0,"reasoning_tokens":2954,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T16:47:43.641540+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a continuous function on the unit disk, vanishing at the origin and satisfying the strong maximum principle, whose zero set intersects each circle centered at the origin at most once (or whose total length is less than 2). For instance, a function with a single smooth zero curve spiraling from the origin to the boundary would, if it existed, refute Theorem 1.9. Alternatively, verifying the covering/rotation inequality (4.13)–(4.21) on a simple piecewise-linear example would test the geometric step.","supporting_citations":[],"review_version":1}