{"id":"7309470a-81a8-4442-8da3-f418d05723e5","arxiv_id":"1908.07122","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Standing waves for NLS with delta-interaction on a star graph are strongly unstable for p at least 5 when alpha is positive, and for p greater than 5 above an explicit frequency threshold when alpha is negative.","lead":"This math paper proves that certain standing waves of the nonlinear Schrödinger equation with a point interaction on a star-shaped graph can be nudged into solutions that blow up in finite time, called strong instability. It also gives similar blow-up theorems for the NLS equation with delta-prime interaction on the line.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.4 rests on inequality (4.13), whose stated verification is missing and whose displayed monotonicity sign is wrong; until (4.13) is independently established, the key bound in Lemma 4.8 is unsupported.","rationale":"I focus on Lemma 4.8 rather than on the externally imported variational identity (3.5). The identity (3.5) is a direct equal-edge reduction to the line results of [13,14], and the scaling constants check out; it is a citation, not an internally deferred proof. By contrast, Lemma 4.8 contains the only step in the main proof that is explicitly left as a verification, and the displayed sentence about the derivative of g has the wrong sign in a concrete case. This is internally testable and directly controls whether B^-_omega is nonempty and invariant. The reader's verdict of CONDITIONAL already accounts for the deferred algebra in Lemma 4.8, so my read does not change the verdict; a filled-in proof of (4.13), or a counterexample, would settle the conditional. I agree with the reader that the variational characterization is load-bearing, but I judge the unverified inequality in Lemma 4.8 to be the least secure point because it is neither proved nor accurately described in the manuscript.","tokens_in":24507,"tokens_out":32503,"duration_ms":310832,"concrete_test":"Independently verify inequality (4.13): for beta > 2 and lambda in (0,1), prove D(lambda) := R2(lambda) - R1(lambda) >= 0, where R1 = beta(lambda^2 + (beta-3)lambda^beta)/(lambda(2-lambda)) and R2 = (2lambda^beta - beta lambda^2 - 2 + beta)/(lambda-1)^2. A concrete first check is to clear denominators and simplify D(lambda)(lambda-1)^2 lambda (2-lambda); then run a high-precision scan over beta in (2,10] and lambda in (0,1) and also check the limiting behavior at lambda -> 0 and lambda -> 1. If D takes a negative value, Lemma 4.8 and Theorem 1.4 fail; if D >= 0 everywhere, the missing proof is fillable and the theorem can stand after the lemma is repaired.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Theorem 1.4 the decisive object is Lemma 4.8, which supplies the bound deq(omega) <= S_omega(V) - P(V)/2 that makes B^-_omega invariant and forces blow-up on that set. The proof of Lemma 4.8 reduces to inequality (4.13) for lambda in (0,1) and beta = (p-1)/2 > 2, but the reduction is not completed. The text says the inequality follows by showing that the derivative of the displayed g is nonpositive; this check is not carried out, and for the displayed g the asserted sign appears wrong near lambda = 1. For example, when beta = 3, the displayed g equals 3lambda/(2-lambda) - (2lambda+1), which increases toward 0 as lambda tends to 1, so its derivative is positive, not nonpositive. The inequality (4.13) may still be true with the opposite monotonicity, but the bridge from (4.6)-(4.12) to (4.13) is missing as written. Since Lemma 4.8 is used in the proof of the invariant set B^-_omega (Lemma 4.10) and hence Theorem 1.4, and since Section 5 states that its delta-prime theorems repeat the proof of Theorem 1.4, the central instability claim for the attractive branch is not fully supported by the manuscript as written. Theorem 1.3 is not affected by this particular gap.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the nonlinear Schr\"odinger equation with a delta interaction on a star graph. It establishes local well-posedness in the operator domain D_H, derives the virial identity for H^1 solutions, and uses a variational characterization of the ground-state profile Phi_0^alpha to prove strong instability by finite-time blow-up: Theorem 1.3 covers the repulsive case alpha > 0 with p >= 5, and Theorem 1.4 covers the attractive case alpha < 0 with p > 5 and omega >= omega_1, where omega_1 is defined by an explicit integral equation. Section 5 states analogous strong instability theorems for the NLS-delta' equation on the line. The proof strategy follows the classical Berestycki-Cazenave method adapted to the graph: invariant sets B_omega^+ and B_omega^- are built from the action S_omega, the virial functional P, and the variational bounds d_eq(omega), and blow-up is forced by the virial identity on these sets.","tokens_in":24834,"tokens_out":5909,"duration_ms":62738,"significance":"If the main theorems are fully proved, the results are a meaningful strengthening of previously known orbital instability results for NLS-delta on star graphs: strong instability by blow-up is a stronger dynamical statement, and the paper supplies a virial identity in a graph setting with a point interaction, which is a useful technical contribution. The explicit threshold omega_1 in Theorem 1.4, obtained from a one-dimensional integral equation, is also a concrete and potentially falsifiable prediction. The paper is transparent about its external inputs, and the well-posedness proof in D_H is given in detail. However, the load-bearing inequality in Lemma 4.8 is not actually verified in the manuscript, and the delta-prime theorems in Section 5 are only sketched, with Remark 5.3 conceding a missing Strichartz ingredient.","major_comments":[{"comment":"The proof of Lemma 4.8 is incomplete. The reduction from (4.6)-(4.12) to inequality (4.13) is not completed, and the suggested verification is incorrect as stated. For beta = 3, the displayed function g satisfies g(lambda) = 3lambda/(2-lambda) - (2lambda+1), whose derivative is 6/(2-lambda)^2 - 2, which is positive near lambda = 1, not nonpositive. The inequality itself appears true in this special case, so a different argument might repair the lemma, but the text provides no such argument. Since Lemma 4.8 is used to prove invariance of B_omega^- in Lemma 4.10 and then Theorem 1.4, the attractive-branch strong instability theorem is not fully supported by the manuscript as written. Theorem 1.3 is not affected by this particular gap.","section":"Section 4.2, Eq. (4.13)"},{"comment":"The proofs of Theorems 5.4 and 5.5 are only presented as 'key steps'. In particular, the virial identity (5.6) is asserted after Remark 5.3, which explicitly concedes that Strichartz estimates for e^{-iH_gamma t} 'might be obtained' using [4] but are not established. Without those estimates, the equality T_{H^1} = T_H used to justify the virial identity for H^1 initial data is not proved. Moreover, the delta-prime analogue of Lemma 4.8 is not stated or proved. Thus, the strong instability results for the delta-prime equation are announced rather than demonstrated in this manuscript.","section":"Section 5, Remark 5.3 and Theorems 5.4-5.5"},{"comment":"The variational identity d_eq(omega) = S_omega(Phi_0^alpha) = N/2 d_line^r(omega) is load-bearing for both main theorems, but it is imported from [13,14] without a proof that the equal-edge lift of the line minimizer remains minimizing among E_eq for the graph. Remark 3.1 explicitly states that for alpha > 0 the variational properties of the profiles were previously unknown. If the cited line results are intended to cover the repulsive branch, the paper should state the precise identification of delta-strengths in the reduction; if not, Theorem 1.3 lacks its key variational input. The same comment applies to the use of [6, Theorem 3.4] for invariance of E_eq under the flow, which is cited without proof.","section":"Section 3, Eq. (3.5) and Remark 3.1"}],"minor_comments":[{"comment":"The displayed integral defining xi_1(p) is ambiguous as printed: it should read ((p-5)/2) integral_xi^1 (1-s^2)^{2/(p-1)} ds = xi(1-xi^2)^{2/(p-1)}.","section":"Theorem 1.4"},{"comment":"The qualitative argument for uniqueness of the zero of f(xi) would be clearer if the derivative f'(xi) were displayed; the current sentence 'f'(xi) has a unique zero' is stated without the formula needed to verify it.","section":"Lemma 4.6"},{"comment":"Remark 4.4 states the alpha = 0 result without proof; a sentence indicating the limiting argument or the needed modification of Lemma 4.1 would be helpful.","section":"Remark 4.4"},{"comment":"The notation phi_omega in Eq. (3.5) depends on both omega and alpha, while phi_{omega,0} is used for alpha=0; this dependence should be made explicit to avoid confusion.","section":"Notation throughout"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about inequality (4.13) lands: the stated derivative check is wrong and the proof of Lemma 4.8 is missing its decisive step. The Section 5 results are also presented at a level more typical of an announcement than a full proof, which is a scope concern for a journal expecting complete arguments. The authors should either supply a correct proof of (4.13) and full details for the delta-prime case, or reframe those results as conditional/sketch."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read of 1908.07122.\n\nThe headline: Theorem 1.3, the repulsive case (alpha > 0, p >= 5), is a clean, substantial contribution and appears to be proved correctly. Theorem 1.4, the main attractive-branch result, has a real gap: the decisive inequality (4.13) in Lemma 4.8 is not proved, and the verification the authors sketch is wrong as stated. They claim the derivative of the displayed g is nonpositive on (0,1), but for beta = 3 the derivative is positive near lambda = 1. The inequality itself may still be true (for beta = 3 it reduces to 3lambda/(2-lambda) <= 2lambda+1, which holds), but the bridge in the manuscript does not establish it. Since Lemma 4.8 is what makes the invariant set B^-_omega work, Theorem 1.4 is not fully supported as written. The delta-prime theorems in Section 5 inherit this problem because their proof is only a key-steps sketch and says to repeat the proof of Theorem 1.4.\n\nWhat is genuinely new and good: the well-posedness theorem in D_H, the virial identity for NLS-delta on a star graph via the H^1 approximation argument, and the blow-up results in regimes not previously covered. The equal-edge reduction d_eq = (N/2) d_line^r is imported from earlier work; that lowers novelty but is not by itself a flaw. The citation pattern is mostly standard for this community, though the paper leans heavily on [6], [13], [14], and [12, Lemma 3.2]. The manuscript is honest about some limitations, e.g. Remark 3.1 concedes that variational properties for alpha > 0 were previously unknown, and Remark 5.3 concedes the Strichartz estimates for the delta-prime flow are not fully supplied.\n\nOne more soft spot worth naming: identity (3.5) is load-bearing for both main theorems, but it is imported rather than reproved. I think the equal-edge scaling is standard and not circular, but given the paper's own Remark 3.1, a complete proof or a clearly stated assumption would close a legitimate loophole.\n\nWho is this for: researchers working on NLS on metric graphs and on point interactions. It deserves a serious referee, but the referee should be asked to verify (4.13) and to fill in the delta-prime analogues before the attractive-branch theorems are accepted. I would not cite Theorem 1.4 in its current form; I would cite the virial identity and the repulsive-branch theorem after checking them. Send it to peer review, not to the desk reject pile.","headline":"Solid repulsive-branch blow-up theorem, but the flagship attractive-branch result rests on an unproved and misdescribed inequality, so the paper deserves a demanding referee rather than immediate acceptance.","tokens_in":25368,"tokens_out":3943,"would_cite":false,"duration_ms":46472,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q55","81Q35","37K40","37K45","47E05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For the NLS equation with δ-interaction on a star graph, the ground-state standing wave is strongly unstable: arbitrarily close initial data blow up in finite time for supercritical nonlinearities.","keywords":["NLS","star graph","delta interaction","standing waves","strong instability","blow-up","virial identity","delta prime interaction"],"falsifier":"Directly compute, for a fixed $\\alpha>0$, $p\\ge 5$, and $\\omega>\\alpha^2/N^2$, the equal-edge Nehari minimum $d_{\\rm eq}(\\omega)$ and compare it with $S_\\omega(\\Phi^\\alpha_0)$; if equality fails for some such parameter triple, Lemma 4.1 and Theorem 1.3 collapse.","tokens_in":24321,"feed_emoji":"💥","tokens_out":5505,"duration_ms":55897,"temperature":0.7,"pith_summary":"The paper studies the nonlinear Schrödinger equation with a δ-interaction at the vertex of a star graph and tries to prove that the ground-state standing waves are strongly unstable, meaning that every neighborhood of the profile contains initial data whose solution blows up in finite time. Two regimes are established: for positive coupling strength $\\alpha>0$ and $p\\ge 5$, and for negative $\\alpha<0$ and $p>5$ above an explicit frequency threshold $\\omega_1$. The proof upgrades earlier orbital-instability results to finite-time blow-up, and it supplies a variational route that avoids detailed spectral analysis. If the claims are correct, the standing wave is not merely unstable in shape but cannot persist dynamically at all in these parameter ranges.","feed_headline":"NLS standing waves on star graphs blow up in finite time","feed_subtitle":"Ground states with δ-coupling are strongly unstable for p≥5, with an explicit threshold when the coupling is negative.","key_machinery":"The load-bearing object is the equal-edge variational identity $d_{\\rm eq}(\\omega)=S_\\omega(\\Phi^\\alpha_0)=\\frac{N}{2}d_{\\rm line}^r(\\omega)$, imported from the line problem, which characterizes the profile as the minimizer of the action $S_\\omega$ on the Nehari manifold inside the equal-edge subspace $E_{\\rm eq}(\\Gamma)$. Around this sit two auxiliary tools: the virial identity $f''(t)=8P(U(t))$ for the second moment $f(t)=\\|xU(t)\\|_2^2$, and the invariance of the sets $B^\\pm_\\omega$ under the flow. The novel variational lemmas (Lemma 4.1 and Lemma 4.8) prove that if $P(V)\\le 0$ and $V$ lies in the relevant class, then $d_{\\rm eq}(\\omega)\\le S_\\omega(V)-\\frac12 P(V)$; this inequality converts the negative curvature of the second moment into a strictly negative bound, yielding finite-time blow-up.","core_discovery":"The central discovery is that the ground-state profile $\\Phi^\\alpha_0$ on an $N$-edge star graph, with equal mass on each edge, is strongly unstable by blow-up for supercritical powers. Theorem 1.3 states that for $\\alpha>0$, $\\omega>\\alpha^2/N^2$, and $p\\ge 5$, the standing wave $e^{i\\omega t}\\Phi^\\alpha_0(x)$ is strongly unstable. Theorem 1.4 states that for $\\alpha<0$, $p>5$, and $\\omega\\ge\\omega_1$, where $\\omega_1=\\alpha^2/(N^2\\xi_1(p)^2)$ and $\\xi_1(p)\\in(0,1)$ is the unique solution of a displayed integral equation, the same strong instability holds. The proof constructs scaled profiles $\\Phi_\\lambda$ arbitrarily close to $\\Phi^\\alpha_0$, shows they enter an invariant set on which the action lies below the ground-state value and the virial functional is negative, and then uses the virial identity to force the second moment of the solution to become negative in finite time, which is impossible unless the solution ceases to exist.","pith_inferences":["A reader might infer that the equal-edge identity $d_{\\rm eq}=N d_{\\rm half}=\\frac{N}{2}d_{\\rm line}^r$ is the real bottleneck: the paper's Remark 3.1 notes that for $\\alpha>0$ the variational properties of the profiles were previously unknown, yet the identity is imported from [13,14] without reproof, so a direct verification for $\\alpha>0$ would remove the main unstated assumption.","One testable extension is to replace the star graph by a finite tree with a δ-coupling at each vertex and check whether the same invariant-set construction forces blow-up on every edge or only on the equal-edge component.","The threshold $\\omega_1$ suggests a sharp stability transition for $\\alpha<0, p>5$: below $\\omega_1$ the proof gives no blow-up, and a companion orbital-stability analysis around $\\omega_1$ would show whether the transition is continuous or discontinuous.","The same variational inequality could be applied to other point-interaction models, such as the δ'-interaction on graphs or networks with different vertex conditions, whenever an equal-edge Nehari minimizer is available."],"forward_implications":["For $\\alpha>0$, the standing wave $e^{i\\omega t}\\Phi^\\alpha_0(x)$ is strongly unstable for all $p\\ge 5$ and all $\\omega>\\alpha^2/N^2$, completing the orbital-instability picture for that branch.","For $\\alpha<0$ and $p>5$, strong instability holds once $\\omega$ exceeds the explicit threshold $\\omega_1$ defined by the unique zero $\\xi_1(p)$ of the displayed integral equation.","In the limit $\\alpha=0$, the same argument gives strong instability of the standard NLS ground state on the star graph for $p\\ge 5$.","The technique transfers to the NLS-$\\delta'$ equation on the line, yielding strong instability of the asymmetric and odd standing waves for $p>5$ above explicit frequency thresholds $\\omega_2$ and $\\omega_3$.","The virial identity and the variational inequality are established for $H^1$-solutions, not only for smooth data, so the blow-up conclusion holds at the natural regularity level of the Cauchy problem."],"supporting_citations":[{"why":"Supplies the line minimizer identity that underpins Eq. (3.5) for the repulsive branch.","marker":"[13]"},{"why":"Provides the variational characterization of the profile for NLS with a point defect on the line, the other anchor of the equal-edge identity.","marker":"[14]"},{"why":"Establishes invariance of the equal-edge subspace $E_{\\rm eq}(\\Gamma)$ under the flow, used in Lemmas 4.2 and 4.10.","marker":"[6]"},{"why":"Supplies the model for the scaling-parameter proof of the key inequality $d_{\\rm eq}\\le S_\\omega(V)-\\frac12 P(V)$ in Lemma 4.8.","marker":"[12]"},{"why":"Provides the variational instability scheme for the NLS equation with a delta potential that Theorems 1.3 and 1.4 adapt to the star graph.","marker":"[19]"},{"why":"Supplies the standard virial-identity and well-posedness machinery used in Propositions 2.5 and Theorem 2.3.","marker":"[11]"},{"why":"Provides the Strichartz estimates used to show that the $D_H$-maximal time and the $H^1$-maximal time coincide.","marker":"[8]"},{"why":"Gives the prior strong-instability result for the NLS-δ equation on the line, the baseline that the star-graph theorem extends.","marker":"[17]"},{"why":"Supplies the variational characterization of the δ'-interaction profiles used in Section 5.","marker":"[1]"}],"fun_headline_variants":["NLS delta-coupling standing waves blow up on star graphs","Star graph standing waves strongly unstable: blow-up for p≥5","Blow-up threshold p≥5 for NLS standing waves on star graphs","Delta-interaction standing waves on star graphs: finite-time blow-up","Strong instability of NLS ground states on star graphs via blow-up"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the equal-edge variational identity $d_{\\rm eq}(\\omega)=S_\\omega(\\Phi^\\alpha_0)=\\frac{N}{2}d_{\\rm line}^r(\\omega)$ holds for both signs of $\\alpha$, together with flow invariance of the equal-edge subspace $E_{\\rm eq}$, and neither fact is reproved in this paper.","fun_headline_variants_meta":{"raw":{"variants":["NLS delta-coupling standing waves blow up on star graphs","Star graph standing waves strongly unstable: blow-up for p≥5","Blow-up threshold p≥5 for NLS standing waves on star graphs","Delta-interaction standing waves on star graphs: finite-time blow-up","Strong instability of NLS ground states on star graphs via blow-up"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001566,"raw_usage":{"total_tokens":6233,"prompt_tokens":902,"completion_tokens":5331,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":518,"completion_tokens_details":{"reasoning_tokens":5239}},"tokens_in":518,"tokens_out":5331,"duration_ms":37506,"temperature":1.0,"reasoning_tokens":5239,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:26:15.661850+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Directly compute, for a fixed $\\alpha>0$, $p\\ge 5$, and $\\omega>\\alpha^2/N^2$, the equal-edge Nehari minimum $d_{\\rm eq}(\\omega)$ and compare it with $S_\\omega(\\Phi^\\alpha_0)$; if equality fails for some such parameter triple, Lemma 4.1 and Theorem 1.3 collapse.","supporting_citations":[{"cited_title":"Fukuizumi, L","cited_arxiv_id":null,"evidence_quote":"Supplies the line minimizer identity that underpins Eq. (3.5) for the repulsive branch."},{"cited_title":"Fukuizumi, M","cited_arxiv_id":null,"evidence_quote":"Provides the variational characterization of the profile for NLS with a point defect on the line, the other anchor of the equal-edge identity."},{"cited_title":"Angulo, N","cited_arxiv_id":null,"evidence_quote":"Establishes invariance of the equal-edge subspace $E_{\\rm eq}(\\Gamma)$ under the flow, used in Lemmas 4.2 and 4.10."},{"cited_title":"Strong instability of standing waves for nonlinear Schr\\\"odinger equations with attractive inverse power potential","cited_arxiv_id":"1804.02127","evidence_quote":"Supplies the model for the scaling-parameter proof of the key inequality $d_{\\rm eq}\\le S_\\omega(V)-\\frac12 P(V)$ in Lemma 4.8."},{"cited_title":"Ohta, Instability of standing waves for nonlinear Schr¨ odinger equations with delta potential, Sao Paulo Journal of Mathematical Sciences 13 (2019), 465–474","cited_arxiv_id":null,"evidence_quote":"Provides the variational instability scheme for the NLS equation with a delta potential that Theorems 1.3 and 1.4 adapt to the star graph."},{"cited_title":"Cazenave, Semilinear Schr¨ odinger equations, Courant Lect","cited_arxiv_id":null,"evidence_quote":"Supplies the standard virial-identity and well-posedness machinery used in Propositions 2.5 and Theorem 2.3."},{"cited_title":"Banica, L.I.Ignat, Dispersion for the Schr¨ odinger equation on the line with mu ltiple Dirac delta potentials and on delta trees , Anal","cited_arxiv_id":null,"evidence_quote":"Provides the Strichartz estimates used to show that the $D_H$-maximal time and the $H^1$-maximal time coincide."},{"cited_title":"Le Coz, R","cited_arxiv_id":null,"evidence_quote":"Gives the prior strong-instability result for the NLS-δ equation on the line, the baseline that the star-graph theorem extends."},{"cited_title":"Adami and D","cited_arxiv_id":null,"evidence_quote":"Supplies the variational characterization of the δ'-interaction profiles used in Section 5."}],"review_version":1}