{"id":"f74bdf4f-2618-408a-a366-614e2368a134","arxiv_id":"1908.03993","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For a nonlinear biharmonic equation on a locally finite graph, ground state solutions exist for any lambda>1, p>2, and as lambda tends to infinity they converge to a ground state solution of the limit problem on the potential well.","lead":"This paper proves existence of ground state solutions for a nonlinear biharmonic equation on locally finite graphs, and shows that as a potential parameter grows, these solutions concentrate on a finite 'potential well' and converge to a solution of a Dirichlet-type limit problem. It extends a known second-order result on graphs to fourth-order equations using variational methods.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified.","rationale":"The reader's CONDITIONAL verdict is reasonable because the manuscript leaves some routine details to the reader, but those details do not threaten the central claim. The weakest assumption identified by the reader, (A2), is indeed the engine of the compact embedding and concentration argument, but it is explicitly assumed and used correctly. The alleged gaps are fillable: Proposition 2.2 is definitional once W^{2,2}(V) is defined as the completion of C_c(V); Theorem 1.2 follows from Theorem 1.3 by constructing the potential a as distance to Ω; and Lemma 3.5 does not actually require density of C_c(V) in Eλ because the same Nehari projection argument works for arbitrary φ∈Eλ. I therefore see no load-bearing correctness risk, though a revision should add these clarifications.","tokens_in":17634,"tokens_out":37338,"duration_ms":385452,"concrete_test":"Re-derive the final step of Lemma 3.5 with an arbitrary φ∈Eλ instead of φ∈C_c(V); if the projection t(s) formula and the differentiation γ'(0) remain valid, the existence theorem holds without extra density assumptions. In parallel, check Theorem 1.2 by verifying that a(x)=dist(x,Ω) satisfies (A1)-(A2) for every bounded connected Ω.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I read the proof as sound in its main line. The central claim requires (i) compact embedding of Eλ into L^p, which Lemma 2.5 establishes using (A2); (ii) a Nehari minimizer that is a critical point in Eλ; and (iii) passage to the limit with u0 supported in Ω. Each of these is justified. The two gaps noted by the reader are presentation issues rather than correctness risks: Proposition 2.2 is essentially a restatement of the definition of W^{2,2}(V) as the completion of C_c(V), and Theorem 1.2 follows from Theorem 1.3 by choosing a(x)=dist(x,Ω) for any bounded connected Ω. The one point that deserves a line in revision is that Lemma 3.5 tests φ∈C_c(V) while Definition 2.1 tests φ∈Eλ; the same proof works verbatim for φ∈Eλ because the projection t(s) formula and the differentiation γ'(0) do not use compact support. I do not find a load-bearing flaw.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the nonlinear biharmonic equation Δ²u − Δu + (λa+1)u = |u|^{p−2}u on a locally finite weighted graph G. Under assumptions (G1)–(G4) and (A1)–(A2), it proves for each λ>1 and p>2 the existence of a ground state solution uλ by Nehari manifold minimization, and then proves that as λ→∞ the ground states converge in W^{2,2}(V) to a ground state solution of the Dirichlet problem Δ²u − Δu + u = |u|^{p−2}u on the potential well Ω={a=0}, with u=0 on ∂Ω. The proof relies on a compact embedding Eλ↪L^q(V) with constants independent of λ, obtained from assumption (A2), and on Nehari-type estimates.","tokens_in":17719,"tokens_out":25634,"duration_ms":238257,"significance":"The result is a natural fourth-order analogue of the deepening-potential-well phenomenon for Schrödinger equations on graphs, extending the second-order work of Zhang and Zhao [31]. The variational framework is standard, but the discrete setting requires careful treatment of Sobolev spaces and integration by parts; the paper provides these tools and a clear proof of the concentration phenomenon. If the gaps noted below are filled, the paper constitutes a solid contribution to the growing literature on nonlinear equations on graphs.","major_comments":[{"comment":"The proof of Lemma 3.5 verifies J'_λ(uλ)φ = 0 only for test functions φ∈C_c(V), while Definition 2.1 requires the identity for all φ∈Eλ. The same projection argument works verbatim for arbitrary φ∈Eλ because the formula for t(s) and the differentiability of γ(s) do not use compact support; please state this explicitly or, alternatively, prove that C_c(V) is dense in Eλ and pass to the limit.","section":"Section 3, Lemma 3.5 and Definition 2.1"},{"comment":"The assertion that u_{λk} → u0 in L^q(V) for all q∈[2,∞) is used to pass to the limit, but Lemma 2.5 is formulated for a fixed λ. A uniform-in-λ compactness argument is needed: using the bound ||u_{λk}||_{E_{λk}} ≤ C (from Remark 4.1 and m_{λk} ≤ mΩ) and assumption (A2), the tail estimate in Lemma 2.5 yields ∫_{d>R}|u_{λk}−u0|²dµ ≤ ε/λ_k for large k, so L² convergence follows; this should be written out or added as a separate lemma.","section":"Section 4, Lemma 4.3 and proof of Theorem 1.3"},{"comment":"The displayed chain contains the equality 'lim inf J_{λk}(t u_{λk}) = ((p−2)/(2p)) lim inf ||t u_{λk}||²_{E_{λk}}'; this is false because t u_{λk} is not generally in N_{λk}. The correct step is '≤', which is sufficient: for t∈(0,1] and u_{λk}∈N_{λk} one has J_{λk}(t u_{λk}) ≥ ((p−2)/(2p))||t u_{λk}||²_{E_{λk}}. The conclusion M ≥ mΩ survives with this correction.","section":"Section 4, Lemma 4.3"}],"minor_comments":[{"comment":"There are typos: 'bihar monic' in the header and 'the the boundary' in the abstract should be corrected.","section":"Title page and abstract"},{"comment":"The repeated 'as k → 0' should be 'as k → ∞', and in the last line of the proof of Proposition 2.1 the limit should be of II_k, not I_k.","section":"Propositions 2.1 and 2.2"},{"comment":"Since W^{2,2}(V) was defined in Section 1 as the completion of C_c(V), Proposition 2.2 is a restatement of the definition, so the omitted boundary-layer estimates for Δ are not a substantive gap; if the authors prefer to define W^{2,2}(V) as the space of functions with finite norm, they should revise the definition and supply the missing estimates.","section":"Proposition 2.2"},{"comment":"In the definition of C1, the expression η_{p+1}^{p+1} should be η_p^p to match the constant used in estimate (17).","section":"Lemma 4.2"},{"comment":"In the proof of strong L² convergence, 'lim inf' should be 'lim'.","section":"Lemma 2.5"},{"comment":"The sentence 'From Lemma 4.1, we have that u0 ≠ 0' is not immediate; one also needs the L^p convergence (19) to rule out u0=0, since Lemma 4.1 only gives a lower bound on the E_{λk}-norms of the u_{λk}.","section":"Lemma 4.3 and proof of Theorem 1.3"},{"comment":"The integration-by-parts lemmas are quoted from [31] without proof; including their statements or a short proof would improve self-containedness.","section":"Lemmas 2.1 and 2.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is a competent extension of [31] to the fourth-order case. The main theorems are plausible and the proof strategy is sound, but the missing arguments for Lemma 3.5 and the uniform-in-λ compactness in Lemma 4.3 need to be supplied before the paper can be accepted. I do not see a fatal flaw; the corrections are local and do not require new ideas."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is the first existence-and-concentration result for a nonlinear biharmonic equation on locally finite graphs, and the core proof holds up. It is a genuine extension of Zhang-Zhao's second-order framework, not a repackaging. The new Sobolev embeddings in Lemma 2.5 are the key added machinery, and the Nehari-manifold argument plus the lambda-to-infinity concentration argument are correctly adapted to the fourth-order discrete setting.\n\nWhat it does well: the paper is honest about borrowing integration-by-parts lemmas from [31], and its own Lemmas 2.3 and 2.4 extend those to fourth order cleanly. The compactness in L^q depends on (A2), and the proof of m_lambda -> m_Omega is careful. I paid special attention to the passage to the limit in Theorem 1.3: the claim that u_0 vanishes outside Omega is justified by the lambda * integral(a |u_lambda|^2) bound, and the proof of (20)-(21) is supported. There is no circularity and no serious self-citation problem.\n\nSoft spots, in proportion:\n- Proposition 2.2 is not really a gap: W^{2,2}(V) is defined as the completion of C_c(V), so the proposition is close to a restatement. The omitted boundary-layer estimate for Delta(eta_k u - u) is said to be similar to the W^{1,2} case; I believe it goes through, but a referee should ask for the display.\n- Theorem 1.2 is derived as a byproduct of Theorem 1.3, not proved directly. To handle an arbitrary bounded connected Omega, one needs a potential a with zero set Omega; the paper does not construct one. A graph analogue of a(x)=dist(x,Omega) does the job, so this is minor.\n- Lemma 3.5 tests only phi in C_c(V) while the weak-solution definition uses phi in E_lambda. As the stress-test note says, the same projection argument works verbatim for phi in E_lambda because the formula for t(s) does not use compact support. One-line fix.\n- Small typos, e.g. 'the the' and a couple of k->0 limits that should be k->infinity. Nothing confusing.\n\nBottom line: the theorems are new and the main estimates are there; the flagged issues are presentation-level, not load-bearing. I would send this to a serious referee. If I worked on graph PDEs I would cite it; it fills a real gap.","headline":"Solid first fourth-order graph analogue of Zhang-Zhao's deepening-well result; the proof's main line is sound and the flagged gaps are presentation-level, so it deserves a serious referee.","tokens_in":18318,"tokens_out":2156,"would_cite":true,"duration_ms":24159,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35A15","35Q55","58E30"],"pacs":[],"model":"deepseek-v4-flash","headline":"On locally finite weighted graphs, a nonlinear biharmonic equation with a deepening potential well admits a ground state for every λ>1, and as λ→∞ the ground states converge to a ground state of the Dirichlet limit problem on the well.","keywords":["biharmonic equation","ground state solution","Nehari manifold","potential well","Sobolev spaces on graphs","locally finite graph","concentration phenomenon"],"falsifier":"Replace (A2) with a bounded potential, for example $a=0$ on a finite interval and $a=1$ elsewhere on an infinite locally finite graph; then $E_\\lambda=W^{2,2}(V)$ and sequences of unit $L^2$ bumps translated to infinity are bounded but have no $L^p$-convergent subsequence, so Lemma 2.5 fails and the stated convergence to the well would require a different compactness argument.","tokens_in":17350,"feed_emoji":"🧮","tokens_out":6841,"duration_ms":65766,"temperature":0.7,"pith_summary":"This paper proves that a fourth-order nonlinear equation on a locally finite weighted graph, $$\\$\\Delta$^2 u - \\$\\Delta$ u + (\\$\\lambda$ a + 1)u = |u|^{p-2}u,$$ with a nonnegative potential $a$ that tends to infinity away from a finite \"potential well\" $\\Omega=\\{a=0\\}$, has a least-energy (ground state) solution for every $\\lambda>1$ and every $p>2$. It then proves the concentration half: as $\\lambda\\to\\infty$, these ground states converge, along subsequences, to a ground state of the Dirichlet boundary-value problem $\\Delta^2 u - \\Delta u + u = |u|^{p-2}u$ in $\\Omega$, $u=0$ on $\\partial\\Omega$. The interest is that this transfers a classical Euclidean phenomenon, deepening potential wells forcing solutions to localize and solve a limit problem on the well, to the discrete graph setting and to fourth-order operators. Along the way the paper supplies the Sobolev embedding facts that make variational methods available on graphs.","feed_headline":"Graph biharmonic equation concentrates into a well","feed_subtitle":"For every lambda>1 a ground state exists; as lambda grows, solutions converge to a Dirichlet limit inside the well.","key_machinery":"The machinery is the weighted graph calculus: the $\\mu$-Laplacian $\\Delta u(x)=\\frac1{\\mu(x)}\\sum_{y\\sim x}\\omega_{xy}(u(y)-u(x))$, the gradient form $\\Gamma(u,v)$, integration-by-parts lemmas for $\\Delta^2$, and the Hilbert spaces $E_\\lambda$ and $H(\\Omega)$ with norms built from $\\Delta$, $\\nabla$, and the potential. The load-bearing device is a compact embedding (Lemma 2.5): $E_\\lambda\\hookrightarrow L^q(V)$ for every $q\\ge2$, uniformly in $\\lambda$, with convergence of bounded sequences pointwise and in every $L^q$. Its proof uses assumption (A2) to control the tail of an $L^2$ difference by the weighted $\\lambda a$ part of the norm. The Nehari manifold $N_\\lambda=\\{u\\ne0:J'_\\lambda(u)u=0\\}$ turns the equation into a minimization problem; on $N_\\lambda$ the identity $\\|u\\|^2_{E_\\lambda}=\\int_V|u|^p\\,d\\mu$ gives the energy formula $J_\\lambda(u)=(\\frac12-\\frac1p)\\|u\\|^2_{E_\\lambda}$. The concentration proof compares $m_\\lambda$ with $m_\\Omega$, uses the tail bound to force the weak limit to vanish outside $\\Omega$, and upgrades weak convergence to strong convergence in $W^{2,2}(V)$.","core_discovery":"The central claim is Theorem 1.1: under graph assumptions (G1)–(G4) and potential assumptions (A1)–(A2), for any $\\lambda>1$, $p>2$, the functional $$J_\\$\\lambda$(u)=\\frac12\\int_V(|\\$\\Delta$ u|^2+|\\nabla u|^2+(\\$\\lambda$ a+1)$u^{2}$)\\,d\\mu-\\frac1p\\int_V |u|^p\\,d\\mu$$ has a minimizer on its Nehari manifold, hence a nontrivial critical point $u_\\lambda$ solving the equation pointwise. Theorem 1.3 asserts that for any sequence $\\lambda_k\\to\\infty$, up to a subsequence $u_{\\lambda_k}\\to u_0$ in $W^{2,2}(V)$, where $u_0$ is a ground state of the limit problem on the potential well with zero boundary data; in particular $u_0$ vanishes outside $\\Omega$. Theorem 1.2, existence for the limit Dirichlet problem, is obtained as a by-product of this convergence rather than by a separate minimization argument. A sympathetic reading is: the fourth-order graph equation inherits both existence and the deepening-well concentration phenomenon from the second-order case, and the discrete structure makes the limit problem finite-dimensional because a bounded domain in a locally finite graph contains finitely many vertices.","pith_inferences":["The same Nehari plus compact-embedding scheme should extend to higher-order polyharmonic equations $\\Delta^m$ on graphs whenever an integration-by-parts identity and a tail estimate of the same type are available.","Because bounded domains in locally finite graphs are finite, the limit problem is finite-dimensional; a direct finite-dimensional minimization should also yield $u_0$, which may simplify concentration proofs in related graph problems.","If $a$ vanishes on several disjoint wells, the argument suggests the limit may select one well according to $p$ and the graph geometry; this selection rule is not investigated in the paper.","One testable modification: replace (A2) by polynomial growth of $a$ along ends and check whether the $\\lambda$-independent tail control survives with a graph-dependent constant; the compact embedding lemma gives a precise condition to verify."],"forward_implications":["The limit problem on any nonempty, connected, bounded domain $\\Omega$ of a locally finite graph has a ground state solution; this is Theorem 1.2, obtained here by letting $\\lambda\\to\\infty$.","Ground states of the penalized equation localize: any weak limit as $\\lambda\\to\\infty$ is identically zero outside the potential well $\\Omega$.","The convergence is strong, not merely weak: $u_{\\lambda_k}\\to u_0$ in $W^{2,2}(V)$, and the energy gap $\\|u_{\\lambda_k}-u_0\\|_{E_{\\lambda_k}}\\to0$.","The embedding $E_\\lambda\\hookrightarrow L^q(V)$ with a $\\lambda$-independent constant supplies a reusable tool for variational problems on locally finite graphs."],"supporting_citations":[{"why":"Supplies the deepening-potential-well convergence method for second-order equations on graphs that this paper extends to the biharmonic case.","marker":"[31]"},{"why":"Establishes the variational framework for nonlinear equations on locally finite graphs that the existence proof builds on.","marker":"[12]"},{"why":"Gives existence of solutions to the Kazdan-Warner equation on finite graph domains, a prototype for the variational methods used here.","marker":"[10]"},{"why":"Provides Yamabe-type equation existence on graphs, supporting the Sobolev-space setup.","marker":"[11]"},{"why":"Studies asymptotic behavior of ground states for nonlinear biharmonic equations on $\\mathbb{R}^N$, the Euclidean counterpart this paper adapts.","marker":"[24]"},{"why":"Develops the deepening potential well concentration idea in the Euclidean setting, motivating the $\\lambda\\to\\infty$ convergence.","marker":"[28]"}],"fun_headline_variants":["Graph biharmonic waves gather in potential well","Limit of graph biharmonic solutions sits in a well","Biharmonic on graphs: from whole graph to a well","As lambda spikes, graph solutions dive into a well"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result rests on the potential growing without bound as the distance from a fixed vertex goes to infinity; if $a$ stays bounded away from the well, the proof's compact embedding fails and the concentration mechanism is not available.","fun_headline_variants_meta":{"raw":{"variants":["Graph biharmonic waves gather in potential well","Limit of graph biharmonic solutions sits in a well","Biharmonic on graphs: from whole graph to a well","As lambda spikes, graph solutions dive into a well"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000563,"raw_usage":{"total_tokens":2710,"prompt_tokens":1025,"completion_tokens":1685,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":641,"completion_tokens_details":{"reasoning_tokens":1621}},"tokens_in":641,"tokens_out":1685,"duration_ms":15037,"temperature":1.0,"reasoning_tokens":1621,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:58:15.558531+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Replace (A2) with a bounded potential, for example $a=0$ on a finite interval and $a=1$ elsewhere on an infinite locally finite graph; then $E_\\lambda=W^{2,2}(V)$ and sequences of unit $L^2$ bumps translated to infinity are bounded but have no $L^p$-convergent subsequence, so Lemma 2.5 fails and the stated convergence to the well would require a different compactness argument.","supporting_citations":[{"cited_title":"Zhang, L","cited_arxiv_id":null,"evidence_quote":"Supplies the deepening-potential-well convergence method for second-order equations on graphs that this paper extends to the biharmonic case."},{"cited_title":"Grigor’yan, Y","cited_arxiv_id":null,"evidence_quote":"Establishes the variational framework for nonlinear equations on locally finite graphs that the existence proof builds on."},{"cited_title":"Grigor’yan, Y","cited_arxiv_id":null,"evidence_quote":"Gives existence of solutions to the Kazdan-Warner equation on finite graph domains, a prototype for the variational methods used here."},{"cited_title":"Grigor’yan, Y","cited_arxiv_id":null,"evidence_quote":"Provides Yamabe-type equation existence on graphs, supporting the Sobolev-space setup."},{"cited_title":"Niu, Z.W","cited_arxiv_id":null,"evidence_quote":"Studies asymptotic behavior of ground states for nonlinear biharmonic equations on $\\mathbb{R}^N$, the Euclidean counterpart this paper adapts."},{"cited_title":"Wang, H.S","cited_arxiv_id":null,"evidence_quote":"Develops the deepening potential well concentration idea in the Euclidean setting, motivating the $\\lambda\\to\\infty$ convergence."}],"review_version":1}