{"id":"024d489e-a2a9-43be-b046-362b9876eaad","arxiv_id":"2412.11822","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"New Pogorelov-type C^2 estimates and rigidity theorems are proved for (k-1)-convex semi-convex solutions of the elliptic and parabolic sum Hessian equations.","lead":"This paper proves interior second derivative estimates for solutions of the sum Hessian equation under a weaker convexity assumption called (k-1)-convexity with an added semi-convexity bound. It then uses these estimates to show that entire solutions with quadratic growth must be quadratic polynomials, and proves parabolic analogues.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.5(a) is false as stated: the lower bound on σ_{k-1} depends on α and K0, so the claimed constant list in Theorems 1.2–1.7 is unsupported.","rationale":"The paper's main contribution is a Pogorelov-type C^2 estimate for (k−1)-convex semi-convex solutions of the sum Hessian equation and the resulting rigidity theorems. The proof follows a known maximum-principle template, and much of the algebra is plausible. The weakest point is not the overall strategy but the auxiliary spectral bounds: Lemma 2.5(a) is used to guarantee a positive lower bound for σ_{k−1} in Γ_{k−1}. That bound is not uniform in α or K0, and the explicit 2D family shows the lemma as stated is false. Since the theorem's constant list omits α and K0, the central claim is stronger than what the proof establishes. This is a conditional gap rather than a collapse: adding α and K0 to the dependencies likely repairs the proof, but the statements must be amended or an additional argument must show the constants can be chosen uniformly. The reader's verdict of CONDITIONAL is therefore appropriate; our check reinforces it but also identifies a concrete false lemma not highlighted by the reader.","tokens_in":29729,"tokens_out":30451,"duration_ms":219465,"concrete_test":"Verify Lemma 2.5(a) with the explicit family above: as α→∞, all hypotheses hold with N0=N1=1 and K0=α^{1/3}, but σ1→0, so no c0(n,k,N0) independent of α exists. Then audit the downstream uses: (i) re-derive Lemma 5.1's estimate with constants allowed to depend on α and K0, and check whether Theorems 1.3/1.7 statements need amendment; (ii) in Lemma 3.2, track the condition λ1≫K0 into the choice of L in (3.25); if L must grow with K0, the final C in (1.4) depends on K0, against the stated dependence.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Lemma 2.5(a) asserts: if N0≤S_k=σk+ασ_{k−1}≤N1 and λn≥−K0, then, for large λ1, σ_{k−1}≥c0(n,k,N0). This is false. Take n=2,k=2 and, for α≫1, λ=(α^{1/3}, −α^{1/3}+α^{−1/3}). Then λ∈Γ1, λ2>−K0 with K0=α^{1/3}, and S_2=σ2+ασ1=(−α^{2/3}+1)+α·α^{−1/3}=1, so N0=N1=1; yet σ1=α^{−1/3}→0. In the proof, the case σk≤0 only yields σ_{k−1}≥N0/α, so the claimed c0 cannot be independent of α (or of K0). This lower bound feeds Lemma 2.5(c), which is used in Lemma 5.1/5.2 for the rigidity theorems; Lemma 3.2 Cases 2–3 also need λ1≫K0 for their error estimates, so the constants in Theorem 1.2 unavoidably inherit K0 unless an additional argument is supplied. As written, the theorem statements listing only n,k,inf f,‖f‖_{C^{1,1}},‖u‖_{C^1},diam are not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Pogorelov-type interior C^2 estimates for the elliptic sum Hessian equation σ_k(D^2u)+ασ_{k-1}(D^2u)=f(x,u,Du) with u=0 on the boundary, under the assumptions that u is (k−1)-convex and semi-convex. It also states parabolic analogues and rigidity theorems for entire solutions with quadratic growth. The elliptic proof follows the maximum-principle template of Li–Ren–Wang, using the test function φ=M log(−u)+log P_m + A/2|Du|^2 + B/2|x|^2, where P_m is built from the shifted eigenvalues κ_i=λ_i+K_0 arising from semi-convexity. The parabolic proof is analogous. Rigidity theorems are obtained by rescaling and applying the new estimates on expanding domains, following the standard Jörgens–Calabi–Pogorelov strategy.","tokens_in":29945,"tokens_out":7783,"duration_ms":73978,"significance":"If the main estimates were valid with the stated dependencies, the paper would give the first Pogorelov-type interior C^2 estimate for the sum Hessian equation in the (k−1)-convex class under semi-convexity, extending earlier results of Liu–Ren and He–Sheng–Xiang–Zhang. The proof is detailed, uses imported structural lemmas from [28,30] and [29], and involves no fitted parameters or circular reasoning. The accompanying Example 1.8, based on Warren's construction, usefully illustrates that the semi-convexity assumption is not redundant. However, the current version contains a load-bearing error in Lemma 2.5(a), so the stated theorems are not established as written.","major_comments":[{"comment":"Lemma 2.5(a) claims a constant c0(n,k,N0) such that σ_{k−1}(λ)≥c0 under N0≤S_k≤N1, λ_n≥−K0, and λ_1 sufficiently large. This is false as stated. For n=k=2 and α≫1, take λ=(α^{1/3}, −α^{1/3}+α^{−1/3}). Then λ∈Γ_1, λ_2≥−K0 with K_0=α^{1/3}, and S_2=σ_2+ασ_1=(−α^{2/3}+1)+α·α^{−1/3}=1, so N_0=N_1=1; yet σ_1=α^{−1/3}→0. The proof's own second case only yields σ_{k−1}≥N_0/α, so c0 cannot be independent of α. Since Lemma 2.5(c), Lemma 5.1, and Lemma 5.2 all rely on this c0, the rigidity arguments in §5 and the constant lists of Theorems 1.2, 1.3, 1.6, and 1.7 are not established as stated.","section":"§2, Lemma 2.5(a)"},{"comment":"The constants in Theorem 1.2 are asserted to depend only on n,k, inf f, ||f||_{C^{1,1}}, ||u||_{C^1}, and diam(Ω). But the proof depends essentially on K_0: the test function is built from κ_i=λ_i+K_0, Lemma 2.5(b) introduces K_0 through |λ_k|≤c0K_0, and Lemma 3.2 Case 3 and Lemma 3.4 use K_0 in essential estimates. Moreover, the parameter α appears in the operator and in Lemma 2.5(a), yet is also omitted. No step in the paper shows that K_0 or α is controlled by the listed quantities. In particular, Lemma 3.2 Case 3 requires λ_1≫K_0 for the error estimates, and the maximum-argument constant L would need to involve K_0. Thus the stated dependence is unsupported even if Lemma 2.5(a) were repaired by allowing α-dependence.","section":"§3, proof of Theorem 1.2"},{"comment":"The final displayed estimate in Lemma 5.1 uses Lemma 2.5(c), namely ∑_i S^{ii}_k ≥ c0 σ_1^{1/(k−2)} for a constant c0 from Lemma 2.5(a). Since Lemma 2.5(a) is false as stated, the conclusion '−C M/u ≥ ... + c0 B/2 u_{11}^{1/(k−2)}' is unjustified. Lemma 5.2 is the parabolic analogue and inherits the same defect. These lemmas are the key step in the rigidity rescaling proofs of Theorems 1.3 and 1.7, so the rigidity theorems are not established by the current argument.","section":"§5, Lemmas 5.1 and 5.2"}],"minor_comments":[{"comment":"The phrase 'there condition positive constants a, b' should read 'there exist positive constants a, b'.","section":"§1, Definition 1.1(3)"},{"comment":"The word 'smei-convexity' is a typo for 'semi-convexity'.","section":"§1, Theorem 1.6"},{"comment":"The sentence 'We can also derive the estimate bu using λ1' contains a typo; it should be 'by using'.","section":"§3, Remark 3.6"},{"comment":"The citation for the parabolic Evans–Krylov regularity theory is given as [33] (Nakamori–Takimoto), while the elliptic version cites [17] (Gilbarg–Trudinger); a standard parabolic reference would be more appropriate and would make the appeal clearer.","section":"§5, proof of Theorem 1.7"},{"comment":"In the sentence 'Based on the analysis in Section 3, we obtain an inequality similar to that in inequality (3.28)', the cross-reference to equation (3.28) is correct but the phrase 'similar to that in inequality (3.28)' is redundant; consider simplifying.","section":"§4.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely salvageable by correcting Lemma 2.5(a) to include α and K_0 in the constants and by restating all theorems with dependence on α and the semi-convexity constant K_0. The novelty claim would then be weaker than advertised but still meaningful. The current version cannot be accepted because the central estimate and the rigidity theorems rely on the false independence claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take on arXiv:2412.11822. The paper extends Pogorelov-type C^2 estimates for the sum Hessian equation σ_k + ασ_{k−1} to (k−1)-convex solutions under a semi-convexity assumption, with parabolic versions and rigidity theorems. The basic idea is sound: use the test function with P_m = Σ(λ_i+K0)^m to handle negative eigenvalues, following Li–Ren–Wang and Tu. That is a legitimate transfer, and the parabolic results plus Example 1.8 are useful.\n\nBut there is a load-bearing gap. Lemma 2.5(a) claims σ_{k−1} ≥ c0(n,k,N0) uniformly when N0 ≤ S_k ≤ N1 and λ_n ≥ −K0. The proof's own case σ_k ≤ 0 only gives σ_{k−1} ≥ N0/α, so the constant depends on α. The stress-test counterexample (n=2,k=2, λ=(α^{1/3}, −α^{1/3}+α^{−1/3})) satisfies the hypotheses with S_2=1 and σ_1→0, so the lemma is false as stated. This is not a minor typo: Lemma 2.5(b) and (c) feed Lemma 3.2, Lemma 3.4, and Lemma 5.1/5.2, so the constants in Theorems 1.2–1.7 unavoidably depend on α and K0. The theorem statements omit both, listing only n,k, inf f, ||f||, ||u||_{C^1}, diam. That means the main theorems are not established as written.\n\nThe fix looks straightforward: add α and K0 to the dependency lists and replace Lemma 2.5(a) with a corrected bound such as c0(n,k,N0,α). The rigidity proof should survive because α and K0 are fixed for a given solution and thus uniform in the rescaling. There are also some sketched “we claim” estimates in Lemma 3.2 and many typos, but nothing suggesting the authors are lost. This is a credible advance if the gaps are repaired.\n\nI’d send it to review rather than desk reject, but with a clear request for full constant tracking and a corrected Lemma 2.5. For a reading group, maybe—it is a good example of how delicate constant dependence is in these maximum-principle arguments. I wouldn’t cite it until the corrected version appears.","headline":"A promising but under-verified extension: the main theorems are not established as stated because Lemma 2.5(a) is false and the constants omit α and K0, though the underlying approach is sound.","tokens_in":30583,"tokens_out":5479,"would_cite":false,"duration_ms":42878,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J60","35K55","35B53","35B45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper establishes Pogorelov-type C^2 estimates and rigidity theorems for (k-1)-convex semi-convex solutions of the sum Hessian equation and its parabolic counterpart.","keywords":["sum Hessian equation","Pogorelov-type C^2 estimate","(k-1)-convex solutions","semi-convexity","rigidity theorem","parabolic Hessian equation","Liouville theorem","fully nonlinear elliptic equations"],"falsifier":"A decisive test is to search for an entire $(k-1)$-convex solution of $\\sigma_k(D^2u)+\\alpha\\sigma_{k-1}(D^2u)=1$ whose Hessian eigenvalues are all bounded below and which satisfies the quadratic growth condition but is not a quadratic polynomial; the existence of such a function would disprove the rigidity theorem. The paper's explicit n=3 example is the nearest candidate, so computing its growth along the third coordinate and its Hessian eigenvalue bounds would settle whether it lies inside or outside the theorem's hypotheses.","tokens_in":29429,"feed_emoji":"📐","tokens_out":11506,"duration_ms":103989,"temperature":0.7,"pith_summary":"This paper proves an interior second-derivative estimate for solutions of the sum Hessian equation $\\sigma_k(D^2u)+\\alpha\\sigma_{k-1}(D^2u)=f(x,u,Du)$ whose Hessian is only $(k-1)$-convex, an assumption weaker than the $k$-convexity needed in earlier work. The extra hypothesis is semi-convexity: all Hessian eigenvalues stay above a fixed negative constant. The estimate has the classical Pogorelov form $(-u)^{\\gamma_0}\\Delta u\\leq C$ and is strong enough to imply that entire solutions on $\\mathbb{R}^n$ with quadratic growth must be quadratic polynomials. The same strategy is carried out for the parabolic equation $-u_t(\\sigma_k+\\alpha\\sigma_{k-1})=f$, giving parabolic Pogorelov estimates and a rigidity theorem $u(x,t)=-mt+p(x)$.","feed_headline":"Weaker convexity still forces entire solutions to be quadratic","feed_subtitle":"Semi-convex (k-1)-convex solutions obey a Pogorelov interior bound; rigidity and parabolic versions follow.","key_machinery":"The load-bearing object is the shifted sum $P_m=\\sum_{i=1}^n\\kappa_i^m$ with $\\kappa_i=\\lambda_i+K_0>0$; it replaces the largest eigenvalue as the quantity whose logarithm is maximized, and its positivity is exactly where semi-convexity enters. Around this object, the proof uses an algebraic lemma stating that when $S_k$ is bounded and $\\lambda_n\\geq -K_0$, the lower-order symmetric functions $\\sigma_{k-1}$ and $\\sigma_{k-2}$ are controlled from below, $|\\lambda_k|\\lesssim K_0$, and $\\sum_i S^{ii}_k$ grows at least like $\\sigma_1^{1/(k-2)}$. These controls let the maximum-principle computation absorb all third-order derivative terms, leaving only $(-u)^{\\gamma_0}\\Delta u$ on one side.","core_discovery":"The central claim is that the sum Hessian operator $S_k=\\sigma_k+\\alpha\\sigma_{k-1}$ still admits a full maximum-principle theory even on the wider $(k-1)$-convex cone, provided the solution is semi-convex. The proof constructs a test function from $P_m=\\sum_i(\\lambda_i+K_0)^m$, where $\\lambda_i$ are the Hessian eigenvalues and $K_0$ is the semi-convexity threshold, and shows that at the maximum point the inequality $(-u)^{\\gamma_0}\\Delta u\\leq C$ holds with constants depending on $n,k,\\inf f,\\|f\\|_{C^{1,1}},\\|u\\|_{C^1}$, and $\\mathrm{diam}(\\Omega)$. Rescaling arguments then convert this estimate into rigidity: entire $(k-1)$-convex semi-convex solutions of $S_k(D^2u)=1$ with quadratic growth are quadratic polynomials, and the parabolic analogue splits as $-mt+p(x)$ with $p$ quadratic.","pith_inferences":["Because the proof uses $K_0$ in essential estimates but the theorem statements do not list it among the constants, the statements implicitly claim uniformity over the semi-convexity threshold; a natural next step is to test whether rescaling the paper's explicit entire solution can make $(-u)^{\\gamma_0}\\Delta u$ blow up while the listed data stay fixed.","The same shifted-sum barrier should extend to longer linear combinations of Hessian elementary functions, such as $\\sigma_k+\\alpha\\sigma_{k-1}+\\beta\\sigma_{k-2}$, and to curvature equations, wherever a lower bound on Hessian eigenvalues is available.","The elliptic theorem leaves the exponent $\\gamma_0$ unspecified while the parabolic gradient-independent theorem gives $1+\\delta$; identifying the optimal exponent in both settings is a natural continuation suggested by the slack in the estimates."],"forward_implications":["For the Dirichlet problem (1.1), every $(k-1)$-convex semi-convex solution has a quantitative interior Hessian bound $(-u)^{\\gamma_0}\\Delta u\\leq C$, which upgrades to interior $C^{2,\\beta}$ estimates by standard regularity theory.","The rigidity theorem shows that any entire $(k-1)$-convex semi-convex solution of $\\sigma_k(D^2u)+\\alpha\\sigma_{k-1}(D^2u)=1$ on $\\mathbb{R}^n$ with quadratic growth is a quadratic polynomial.","For the parabolic equation (1.7), $(k-1)$-convex-monotone solutions satisfy $(-u)^{1+\\delta}\\Delta u\\leq C$ when $f$ is independent of the gradient, and $(-u)^{\\gamma_0}\\Delta u\\leq C$ when $f$ depends on the gradient and $u$ is semi-convex.","The parabolic rigidity theorem says that entire solutions of $-u_t(\\sigma_k+\\alpha\\sigma_{k-1})=1$ with quadratic growth at $t=0$ and $0<m_1\\leq -u_t\\leq m_2$ have the form $u(x,t)=-mt+p(x)$ with $m>0$ and $p$ a quadratic polynomial.","The example in the paper shows that some extra assumption beyond $(k-1)$-convexity is needed for rigidity, since there exist non-polynomial entire $(k-1)$-convex solutions of the same equation."],"supporting_citations":[{"why":"Supplies the k-convex Pogorelov estimate for the sum Hessian equation and the key lemma that this paper weakens to (k-1)-convexity.","marker":"[30]"},{"why":"Establishes parabolic Pogorelov estimates for (k+1)-convex-monotone k-Hessian solutions, the result here extended to the sum Hessian case.","marker":"[25]"},{"why":"Defines the admissible cone for $\\sigma_k+\\alpha\\sigma_{k-1}$ and the ellipticity framework used throughout the maximum-principle proof.","marker":"[28]"},{"why":"Supplies the third-order cancellation technique and the rescaling rigidity argument that the test-function computation adapts.","marker":"[29]"},{"why":"Gives the non-polynomial entire solution construction used in Example 1.8 to show the rigidity theorem's assumptions are not redundant.","marker":"[44]"},{"why":"Provides the parabolic k-Hessian interior-estimate structure that Theorems 1.5 and 1.6 follow.","marker":"[4]"},{"why":"Introduces semi-convexity into Pogorelov estimates for Hessian equations, the assumption adopted here for the sum Hessian case.","marker":"[43]"},{"why":"Supplies the eigenvalue-derivative lemma used in the parabolic maximum-principle computation.","marker":"[6]"}],"fun_headline_variants":["Semi-convexity still enforces sum Hessian rigidity","Pogorelov estimates hold for semi-convex (k-1)-convex","Quadratic endgame: semi-convex global solutions","Parabolic sum Hessian inherits rigidity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on semi-convexity — every Hessian eigenvalue exceeds a fixed negative constant $-K_0$ — because the barrier $P_m=\\sum(\\lambda_i+K_0)^m$ needs its shifted eigenvalues positive; the theorem's constant list does not include $K_0$, so the statement also implicitly claims the estimate is uniform in that threshold.","fun_headline_variants_meta":{"raw":{"variants":["Semi-convexity still enforces sum Hessian rigidity","Pogorelov estimates hold for semi-convex (k-1)-convex","Quadratic endgame: semi-convex global solutions","Parabolic sum Hessian inherits rigidity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000301,"raw_usage":{"total_tokens":1690,"prompt_tokens":856,"completion_tokens":834,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":472,"completion_tokens_details":{"reasoning_tokens":764}},"tokens_in":472,"tokens_out":834,"duration_ms":7791,"temperature":1.0,"reasoning_tokens":764,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T14:32:46.844050+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test is to search for an entire $(k-1)$-convex solution of $\\sigma_k(D^2u)+\\alpha\\sigma_{k-1}(D^2u)=1$ whose Hessian eigenvalues are all bounded below and which satisfies the quadratic growth condition but is not a quadratic polynomial; the existence of such a function would disprove the rigidity theorem. The paper's explicit n=3 example is the nearest candidate, so computing its growth along the third coordinate and its Hessian eigenvalue bounds would settle whether it lies inside or outside the theorem's hypotheses.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the k-convex Pogorelov estimate for the sum Hessian equation and the key lemma that this paper weakens to (k-1)-convexity."},{"cited_title":"He, H Sheng, N","cited_arxiv_id":null,"evidence_quote":"Establishes parabolic Pogorelov estimates for (k+1)-convex-monotone k-Hessian solutions, the result here extended to the sum Hessian case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the admissible cone for $\\sigma_k+\\alpha\\sigma_{k-1}$ and the ellipticity framework used throughout the maximum-principle proof."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the third-order cancellation technique and the rescaling rigidity argument that the test-function computation adapts."},{"cited_title":"Warren, Nonpolynomial entire solutions to σk equations","cited_arxiv_id":null,"evidence_quote":"Gives the non-polynomial entire solution construction used in Example 1.8 to show the rigidity theorem's assumptions are not redundant."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the parabolic k-Hessian interior-estimate structure that Theorems 1.5 and 1.6 follow."},{"cited_title":"Pogorelov type estimates for $(n-1)$-Hessian equations and related rigidity theorems","cited_arxiv_id":"2405.02939","evidence_quote":"Introduces semi-convexity into Pogorelov estimates for Hessian equations, the assumption adopted here for the sum Hessian case."},{"cited_title":"Brendle, K","cited_arxiv_id":null,"evidence_quote":"Supplies the eigenvalue-derivative lemma used in the parabolic maximum-principle computation."}],"review_version":1}