{"id":"d647a0c2-8d94-4655-bac3-9b9e503e5bda","arxiv_id":"2608.04422","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For sigma_k(-D^2u)=u^p in R^n, the paper proves all nonnegative entire solutions vanish for the previously open exponent range, and identifies the critical exponent as the sharp Liouville threshold.","lead":"The paper proves a sharp no-solution theorem for a family of fully nonlinear PDEs, the k-Hessian Lane-Emden equations, in the previously open exponent range, and classifies the special critical solutions that do exist. It closes a twenty-year gap and gives a fully nonlinear analogue of classical Liouville and classification results for the Laplacian.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bubble normalization (1.9) in Theorem 1.5 is algebraically wrong (n/k should be binom(n,k)), and the n=4k weighted function h in (5.18) is incompatible with the corrected bubble, so Proposition 5.5 and the n=4k classification are not established as written.","rationale":"I examined the proof of Lemma 2.4 and found the quantitative Newton-Maclaurin inequality to be internally sound: the diagonal identity (2.12) checks out, and the two parentheses are nonnegative by the stated Newton-Maclaurin applications. So the reader's weakest assumption did not reveal a concrete gap. The actual soft spot is in the algebra of the critical case: the normalization (1.9) is demonstrably inconsistent with direct differentiation of the bubble and with the paper's own Remark 1.6. This is not a cosmetic issue, because the converse assertion of Theorem 1.5 describes the wrong set of functions, and because the n=4k weighted-function construction (5.18)-(5.24) uses this normalization and produces an equation for h that the standard bubble does not satisfy. Proposition 5.5 is the load-bearing step for divJ identically zero at n=4k, and as written its proof relies on the incorrect (5.24). The central subcritical Liouville theorem (Theorem 1.1) may be unaffected, and the rigidity mechanism for 2k<n<4k may also survive, but the critical classification at n=4k and the exact statement of the bubble family need correction. I therefore recommend CONDITIONAL rather than unconditional acceptance: the authors should fix the normalization to binom(n,k), re-derive the h-definition, and verify Proposition 5.5 with the corrected constants.","tokens_in":56507,"tokens_out":24049,"duration_ms":209508,"concrete_test":"For k=2, n=8, substitute u=(a+b|x-x0|^2)^{-1} into sigma_2(-D^2u)=u^5 at x=x0. Direct computation gives sigma_2 = 112 b^2 a^{-4} and u^5 = a^{-5}, hence 112 a b^2 = 1, in contradiction to (1.9) which requires 16 a b^2 = 1. Separately, plug the correctly normalized bubble into (5.18) with h=b0: the left side equals binom(8,2)^{2} * 4 a b0^2 = 28, not 1. If these checks reproduce the mismatch, the normalization (1.9) and the defining equation (5.18) must be corrected, and Proposition 5.5 must be re-verified with the corrected h before Theorem 1.5 can be accepted at n=4k.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The critical classification theorem is stated with the bubble normalization (1.9): (n/k)a0((n-2k)/k b0)^k = 1. Direct differentiation of u=(a0+b0|x-x0|^2)^{-(n-2k)/(2k)} at x0 gives -D^2u(x0)=((n-2k)/k)b0 a0^{-n/(2k)} Id, hence sigma_k(-D^2u(x0)) = binom(n,k)((n-2k)/k b0)^k a0^{-n/2}. The equation sigma_k = u^{p*} gives a0^{-(n+2)/2} on the right, so the correct condition is binom(n,k) a0 ((n-2k)/k b0)^k = 1. The factor n/k in (1.9) is wrong; for k=2,n=8 it gives ab^2=1/16, while the paper's own Remark 1.6 requires ab^2=1/112. Because of this error, the 'Conversely' part of Theorem 1.5 is false as stated. The error propagates into the n=4k cutoff argument: with the correctly normalized bubble, h=b0 does not satisfy (5.18). For k=2, the left side of (5.18) equals 28^2 * 4ab^2 = 28, not 1, so the claimed h=b0 for the standard bubble fails. Moreover the normalized equation (5.24), whose constant is 2k binom(4k-1,k-1), does not follow from (5.19); (5.19) would instead yield binom(4k,k)^{2k-2} times the bracket. Proposition 5.5, which is the sole tool producing div(h^{-epsilon}J) >= (1/2)h^{-epsilon} divJ and hence divJ identically zero at n=4k, relies on (5.24) and is therefore not proved by the text as it stands. The rigidity conclusion that u is a bubble may survive with corrected constants, but the n=4k case of Theorem 1.5 is not currently supported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies nonnegative entire solutions of the k-Hessian Lane–Emden equation σ_k(−D^2u)=u^p in R^n with −D^2u∈Γ_k and 2≤k<n/2. The main results are: (i) an unconditional Liouville theorem in the subcritical range p_−<p<p_* for C^2 solutions (Theorem 1.1), extended to locally bounded Hessian-measure weak solutions (Theorem 1.2); (ii) classification at the critical exponent p=p_*: for 2k<n≤4k every nonnegative C^2 solution is a Hessian–Sobolev bubble (Theorem 1.5), with conditional statements for n>4k and an exponential counterpart at n=2k (Theorem 1.12). The proof introduces a quantitative Newton–Maclaurin matrix inequality (Lemma 2.4) with an equality case (Lemma 2.5), a vector field J whose divergence is coercive in the subcritical case and nonnegative in the critical case, and weighted cutoff estimates. The paper also contains an appendix justifying the distributional divergence calculations for C^2 solutions.","tokens_in":56964,"tokens_out":13769,"duration_ms":125884,"significance":"If the main results are correct, Theorem 1.1 closes a gap left open by Phuc–Verbitsky and Ou and identifies p_* as the sharp Liouville threshold for the k-Hessian equation; the critical classification extends the Caffarelli–Gidas–Spruck theorem to fully nonlinear Hessian equations. The matrix inequality (2.11) and its equality characterization are elegant and likely to be useful beyond this paper. The weak-solution extension and the careful C^2 justification are additional strengths. However, the critical classification contains algebraic errors in the bubble normalization and in the n=4k weighted argument. These errors affect the statement of Theorem 1.5 and the proof of the n=4k case, so the advertised classification is not established as written; the subcritical Liouville theorems appear independent of these specific constants and are plausibly sound.","major_comments":[{"comment":"The normalization (1.9) is algebraically wrong. Direct differentiation at x=x0 gives −D^2u(x0)=((n−2k)/k)b0 a0^{−n/(2k)} Id, hence σ_k(−D^2u(x0))=binom(n,k)((n−2k)/k b0)^k a0^{−n/2}. Since u(x0)^{p_*}=a0^{−(n+2)/2}, the equation forces binom(n,k) a0 ((n−2k)/k b0)^k = 1. The factor n/k in (1.9) is incorrect for k≥2; for k=2,n=8 it gives a0 b0^2=1/16, whereas Remark 1.6 requires a0 b0^2=1/112. Consequently the 'Conversely' assertion in Theorem 1.5 is false as stated, and Theorems 1.7, 1.8, and Corollary 1.10 inherit the wrong normalization condition in their conclusions.","section":"Theorem 1.5, Eq. (1.9)"},{"comment":"The n=4k weighted construction is incompatible with the standard bubble. At n=4k the bubble is w=u^{−1}=a0+b0|x−x0|^2, and the text states before (5.18) that one wants h=b0 for this function. Substituting h=b0 into (5.18) gives 4 binom(4k,k)^{2k−2} a0 b0^k = 1, while the corrected equation normalization gives a0 b0^k = 1/(2^k binom(4k,k)); for k=2 the left side equals 28, not 1, so h=b0 does not solve (5.18) for the actual bubble. Moreover, after the normalization u^2h=1, equation (5.19) yields u^{2k+1}=binom(4k,k)^{2k−2}(4−|∇u|^2/u), not the expression 2k binom(4k−1,k−1)(4−|∇u|^2/u) stated in (5.24). Since Proposition 5.5 and the subsequent inequality div(h^{−ε}J)≥(1/2)h^{−ε} divJ rely on these identities, the n=4k case of Theorem 1.5 is not proved as written.","section":"Eqs. (5.18), (5.19), (5.24); Proposition 5.5"}],"minor_comments":[{"comment":"The paragraph beginning 'The rest of the paper is organized as follows' appears twice, verbatim, before Section 2; one copy should be deleted.","section":"Section 1.5"},{"comment":"The sentence 'Thenumust be the Aubin-Talenti type bubble' contains a typo and should read 'Then u must be the Aubin-Talenti type bubble'.","section":"Corollary 1.13"},{"comment":"The remark says that the rearrangement for −1<δ<0 'does not follow from a one-sided Hessian inequality'; this is helpful, but the phrase 'the rearrangement uses the equality σ_k(A)=u^p' could be expanded to clarify that the sign change in b_s is the reason for the rearrangement.","section":"Remark 3.2"}],"recommendation":"major_revision","confidential_remarks":"The algebraic errors in the critical classification are serious but appear to be localized: the normalization (1.9) is fixed by replacing n/k with binom(n,k), and the n=4k weighting (5.18)–(5.24) needs a corrected constant and a reworked Proposition 5.5. The subcritical Liouville theorems, which are the paper's first main contribution, seem independent of these issues. I recommend major revision rather than rejection, but the authors must either repair the n=4k argument or restrict Theorem 1.5 to 2k<n<4k with a clearly stated open problem for n=4k."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is a substantial paper with a real subcritical theorem and a load-bearing error in the critical classification. The normalization in Theorem 1.5 is wrong, and the n=4k case is not proved as written. It should go to referees, but it needs major revision.\n\nWhat is actually new: the subcritical Liouville theorem closing (p_-, p_*) is the first complete result in that range, and the vector field J plus the quantitative Newton–Maclaurin inequality (2.11) are genuine new tools. The weak-solution theorem and the approximation argument look credible. I did not find problems in the subcritical cutoff estimates or in Lemma 2.4; those parts read carefully and honestly.\n\nThe soft spot is concrete. Direct differentiation of the bubble at x0 gives the normalization condition binom(n,k) a0 ((n-2k)/k b0)^k = 1, not (n/k) a0 ((n-2k)/k b0)^k = 1. The paper's own Remark 1.6 for k=2 is consistent with the binomial constant and contradicts (1.9). So the converse part of Theorem 1.5 is false as stated.\n\nThe n=4k argument inherits this. With the correct bubble normalization, the function h defined by (5.18) is not identically b0; for k=2 it is b0/28. Then (5.24) does not follow from (5.19): the left side is \\binom{4k}{k}^{2k-2}(4-|\\nabla u|^2/u), not 2k\\binom{4k-1}{k-1}(...). Proposition 5.5, which is the only step producing div(h^{-\\epsilon}J) >= ... and hence divJ = 0 at n=4k, is not established by the text. The classification for 2k<n<4k may survive with the corrected constant, but the n=4k case is currently unsupported.\n\nRecommendation: send to peer review; the subcritical part deserves serious attention and the critical part may be fixable. Do not accept the current version.","headline":"The subcritical Liouville theorem is a strong twenty-year gap closure, but the critical classification has a wrong normalization and the n=4k case is unsupported as written; needs major revision before acceptance.","tokens_in":57527,"tokens_out":8046,"would_cite":false,"duration_ms":75581,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J60","35B53","35B33"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the exponent $p_* = k(n+2)/(n-2k)$ is the sharp Liouville threshold for the $k$-Hessian Lane–Emden equation and classifies critical solutions as explicit bubbles for $2k < n \\le 4k$.","keywords":["k-Hessian equation","Lane–Emden equation","Liouville theorem","critical exponent","critical bubble","Newton tensor","Hessian measure","rigidity"],"falsifier":"A concrete check is to sample eigenvalues in $\\Gamma_k$ and compute the smallest eigenvalue of $\\frac{n-k}{n}\\operatorname{tr}(L_k(A)A)T_{k-1}(A)-L_k(A)^2$; any negative value would invalidate the central rigidity step. Alternatively, numerically integrate the radial $k$-Hessian equation for $k=2$, $n=5$, and $p$ just below $p_*=14/3$; a bounded positive admissible solution would disprove the Liouville theorem.","tokens_in":56294,"feed_emoji":"🧮","tokens_out":13033,"duration_ms":122536,"temperature":0.7,"pith_summary":"The paper studies nonnegative solutions of the fully nonlinear equation $\\sigma_k(-D^2u)=u^p$ in $\\mathbb{R}^n$ that are admissible, meaning $-D^2u$ lies in the cone $\\Gamma_k$ where the first $k$ elementary symmetric functions of the Hessian are nonnegative. It proves that for $2\\le k<n/2$, no nontrivial solution exists when $p$ lies strictly between $p_- = nk/(n-2k)$ and the critical exponent $p_* = k(n+2)/(n-2k)$. Combined with earlier nonexistence results below $p_-$ and known radial solutions for $p\\ge p_*$, this identifies $p_*$ as the sharp Liouville threshold. At $p=p_*$, the paper shows that for $2k<n\\le 4k$ every nonnegative $C^2$ solution is either zero or an explicit critical bubble, with no extra assumptions at infinity. The same Liouville conclusion is extended to locally bounded Hessian-measure weak solutions, and the limiting dimension $n=2k$ is handled for the exponential equation $\\sigma_k(-D^2u)=e^u$ under an integral growth condition.","feed_headline":"Critical exponent proven sharp for k-Hessian Lane-Emden equation","feed_subtitle":"Positive solutions vanish below p*, and critical solutions are explicit bubbles in low dimensions.","key_machinery":"The load-bearing object is the vector field $J$ together with the matrix inequality (2.11). For $A\\in\\Gamma_k$, define $T_j(A)$ by the Newton-tensor expansion and $L_k(A)=\\frac{n-k}{n}\\sigma_k(A)I-T_k(A)$; the inequality $L_k(A)^2\\preceq \\frac{n-k}{n}\\operatorname{tr}(L_k(A)A)T_{k-1}(A)$ is a quantitative Newton–Maclaurin estimate whose equality case (Lemma 2.5) forces all eigenvalues of the deleted tuple to coincide. The divergence of $J$ turns this linear-algebraic control into pointwise coercivity of weighted gradients, and in the critical case into nonnegativity whose vanishing implies the Hessian of $u^{-2k/(n-2k)}$ is a scalar matrix. In the endpoint dimension $n=4k$, the plain vector field degenerates and the proof replaces $J$ by $h^{-\\varepsilon}J$, where $h$ solves an algebraic equation depending on $u$ and $|\\nabla u|^2$; this restores the coercivity needed to conclude $\\operatorname{div}J\\equiv 0$.","core_discovery":"The central discovery is that a single divergence identity for the vector field $J=-u^M L_k(A)\\nabla u-\\tau u^{M-1}|\\nabla u|^2 T_{k-1}(A)\\nabla u$, with $A=-D^2u$, separates nonexistence from classification. The trace-free tensor $L_k(A)=\\frac{n-k}{n}\\sigma_k(A)I - T_k(A)$ and the quantitative Newton–Maclaurin inequality $L_k(A)^2 \\preceq \\frac{n-k}{n}\\operatorname{tr}(L_k(A)A)T_{k-1}(A)$ control the mixed terms: in the subcritical range the divergence of $J$ dominates $u^{M+p-1}|\\nabla u|^2$, so cutoff estimates force every integral of a positive power of $u$ to vanish; at the critical exponent the divergence is nonnegative and the cutoff argument forces $\\operatorname{div}J\\equiv 0$. The equality case of the matrix inequality then yields $D^2(u^{-2k/(n-2k)})=\\lambda\\,\\mathrm{Id}$, so $u(x)=(a_0+b_0|x-x_0|^2)^{-(n-2k)/(2k)}$ with the normalization $\\frac{n}{k}a_0\\left(\\frac{n-2k}{k}b_0\\right)^k=1$. This is the fully nonlinear counterpart of the classical semilinear Liouville and classification theorems, with the same bubble family attaining equality in the sharp Hessian energy inequality.","pith_inferences":["The proof suggests that the same matrix inequality should yield a quantitative stability theorem for the Hessian energy inequality: an appropriately Newton-tensor-weighted distance to the bubble family should be controlled by the energy deficit.","The dimensional line $2k<n\\le 4k$ is where the cutoff powers have favorable signs; the weighted-vector-field trick used at $n=4k$ is a natural template for removing the growth and boundedness assumptions in $n>4k$, so those extra hypotheses may be technical.","A testable intermediate step is whether bounded solutions already satisfy the unconditional classification in dimensions $n=4k+3$ and higher, which would confirm that the remaining gap is a proof-technicality rather than a true phenomenon."],"forward_implications":["The critical exponent $p_*$ is sharp: no nonnegative admissible solution exists for $0<p<p_*$, while radial positive solutions exist for every $p\\ge p_*$.","For $2k<n\\le 4k$ and $p=p_*$, every positive $C^2$ solution is a critical bubble, so the extremals of the sharp Hessian energy inequality are fully classified.","The same Liouville threshold holds for locally bounded weak solutions in the Hessian-measure sense, so the result does not depend on extra $C^2$ regularity.","For $n=2k$, every solution of the exponential equation satisfying the integral growth condition (1.15) is a logarithmic bubble, and its total mass is computed explicitly.","The Liouville theorem supplies universal boundary blow-up estimates of the form $\\sup_\\Omega \\operatorname{dist}(x,\\partial\\Omega)\\,u(x)^{(p-k)/(2k)}\\le C$ for admissible solutions on bounded domains."],"supporting_citations":[{"why":"Supplies the prior nonexistence theorem for the lower range $k<p\\le p_-$, so the new result completes the full subcritical interval.","marker":"[60]"},{"why":"Supplies the prior nonexistence theorem for $0<p\\le k$, covering the complementary edge of the gap.","marker":"[56]"},{"why":"Supplies radial existence of positive solutions for $p\\ge p_*$ and the radial bubble formula, making the threshold sharp.","marker":"[84]"},{"why":"Supplies the potential estimate used to regularize and prove positivity of Hessian-measure weak solutions.","marker":"[44]"},{"why":"Supplies the Newton-tensor identities, Newton–Maclaurin inequalities, and $k$-convex regularity tools used throughout the proof.","marker":"[83]"},{"why":"Supplies the Hessian-measure weak formulation, weak continuity, and comparison principles needed for the weak Liouville theorem.","marker":"[72, 73, 74]"}],"fun_headline_variants":["Critical exponent p* sharp for k-Hessian Lane-Emden","Optimal Liouville theorem proved below critical p*","Sharp threshold: k-Hessian solutions vanish below p*","Bubble classification at critical Hessian exponent","Fully nonlinear Liouville theorem: critical p* proven"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof depends on a matrix inequality proved inside the paper—that the square of the trace-free tensor $L_k(A)$ is bounded by $\\frac{n-k}{n}\\operatorname{tr}(L_k(A)A)T_{k-1}(A)$ for every admissible Hessian $A$—together with its equality characterization; if either the inequality or the equality case had any exception, both the nonexistence theorem and the bubble classification would fail.","fun_headline_variants_meta":{"raw":{"variants":["Critical exponent p* sharp for k-Hessian Lane-Emden","Optimal Liouville theorem proved below critical p*","Sharp threshold: k-Hessian solutions vanish below p*","Bubble classification at critical Hessian exponent","Fully nonlinear Liouville theorem: critical p* proven"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000676,"raw_usage":{"total_tokens":3219,"prompt_tokens":1235,"completion_tokens":1984,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":851,"completion_tokens_details":{"reasoning_tokens":1919}},"tokens_in":851,"tokens_out":1984,"duration_ms":16638,"temperature":1.0,"reasoning_tokens":1919,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:17:42.333872+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check is to sample eigenvalues in $\\Gamma_k$ and compute the smallest eigenvalue of $\\frac{n-k}{n}\\operatorname{tr}(L_k(A)A)T_{k-1}(A)-L_k(A)^2$; any negative value would invalidate the central rigidity step. Alternatively, numerically integrate the radial $k$-Hessian equation for $k=2$, $n=5$, and $p$ just below $p_*=14/3$; a bounded positive admissible solution would disprove the Liouville theorem.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the prior nonexistence theorem for the lower range $k<p\\le p_-$, so the new result completes the full subcritical interval."},{"cited_title":"Ou, Nonexistence results for Hessian inequality,Methods Appl","cited_arxiv_id":null,"evidence_quote":"Supplies the prior nonexistence theorem for $0<p\\le k$, covering the complementary edge of the gap."},{"cited_title":"Wang and Y","cited_arxiv_id":null,"evidence_quote":"Supplies radial existence of positive solutions for $p\\ge p_*$ and the radial bubble formula, making the threshold sharp."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the potential estimate used to regularize and prove positivity of Hessian-measure weak solutions."},{"cited_title":"Wang, Thek-Hessian equation, inGeometric Analysis and PDEs, Lecture Notes in Math.1977, Springer, Dordrecht, 2009, 177–252","cited_arxiv_id":null,"evidence_quote":"Supplies the Newton-tensor identities, Newton–Maclaurin inequalities, and $k$-convex regularity tools used throughout the proof."}],"review_version":1}