{"id":"7b878a4d-bbe5-4f12-b74d-4b9b690f2bdc","arxiv_id":"1908.01532","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For a one-parameter family of tronquee Painleve II solutions, explicit asymptotic constants for the total integrals of the solution and of the associated Hamiltonian are derived for all alpha > -1/2 and omega >= 0.","lead":"This paper derives precise asymptotic formulas, including constant terms, for regularized integrals of a family of Painleve II solutions and their Hamiltonians. The formulas are relevant because such integrals control large-gap tail probabilities in random matrix theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proof of Theorem 1.3 integrates asymptotic identities in α and β without uniform error bounds, and eq. (4.23) appears to misstate a factor; the constant terms in (1.34)–(1.35) need a repaired derivation.","rationale":"The reader's weakest-assumption diagnosis is essentially correct: the proof of Theorem 1.3 integrates asymptotic relations over α and β without an explicit uniformity argument for the error terms. My own reading confirms this gap, and I also found a concrete algebraic symptom in (4.23), where the derivative of w appears to be misstated by a factor of sec^2 B (equivalently, the first line should contain cos(A)/cos(B), not cos(A)cos(B)). This is not an objection to the final formulas: the special cases α=0 and ω=0 match known results, and the corrected ratio identity leads to the same second line used in the rest of the proof. The issue is therefore a repairable rigor gap rather than a demonstrated counterexample. Since the reader already assigned CONDITIONAL for essentially this reason, my concern does not move the verdict; it does sharpen the required revision: either add uniform parameter-dependent error estimates or explicitly justify differentiating the oscillatory asymptotics before integrating them.","tokens_in":37434,"tokens_out":21999,"duration_ms":220120,"concrete_test":"Re-derive (4.23) from (3.75)–(3.76) with an explicit uniform remainder bound on compact parameter intervals, correctly accounting for the derivative of the O(s^{-1}) term; verify that the corrected identity is 2|α−β| cos(A)/cos(B) times dαB. Then substitute the corrected identity into (4.5) and repeat the integrations leading to (4.14), (4.24), and (4.25), tracking the integrated remainder. If the integrated remainder is o(1) as |s|→∞, the constants in (1.34)–(1.35) are fixed; if it is only O(1), the theorem's stated remainder is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claims (1.20)–(1.21) and (1.34)–(1.35) rest on Lemma 4.1 and the integrations in Section 4.3. In particular, (4.5), (4.10), (4.14)–(4.15), and (4.21)–(4.25) differentiate I3 with respect to α and β and then integrate over parameter intervals. The O(|s|^{-3/2}) remainders in Propositions 3.1–3.4 are stated only for fixed α and ω; no uniform-in-parameter version is proved. This matters because ∂αw enters through the oscillatory asymptotics (3.75)–(3.76), and differentiating the O(s^{-1}) remainder in (3.76) is not justified by the stated asymptotics alone. Without such control, the passage from (4.14) to (4.15) and from (4.23)–(4.24) to (4.25) is not fully established, so the remainders claimed in Theorem 1.3 are not proven. A second, concrete symptom appears in (4.23): differentiating w = |s|^{1/2} tan B + O(s^{-1}) introduces a factor sec^2 B, so the first displayed expression should be 2|α−β| cos(A)/cos(B) times dαB, not 2|α−β| cos(A)cos(B) times dαB. The later line (2iβ − 2α tanB)dαB is consistent with the ratio form, but the printed derivation needs correction before the integration step is valid.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a family of tronquée solutions of the Painlevé II equation with parameter ν=2α+1/2 and Stokes multipliers (1.14), for α>-1/2 and ω≥0. It defines regularized integrals I1 of the solution q and I2 of the associated Hamiltonian H, and establishes explicit asymptotic expansions as s→+∞ (Theorem 1.1) and as s→−∞ (Theorem 1.3), with all constant terms evaluated in terms of the Gamma function, Barnes G-function, the derivative of the Riemann zeta function, and arg Gamma. The method combines the Riemann-Hilbert analysis for the Painlevé XXXIV equation from earlier work of the authors and others with differential identities (Lemma 2.3 and Lemma 4.1) that relate the regularized integrals to the asymptotics of the Hamiltonian and related quantities. Special cases reproduce previously known results, including the Hastings-McLeod total integral (1.5), the α=0, ω=1 special solution, and the large-gap expansion of Bogatskiy-Claeys-Its.","tokens_in":37753,"tokens_out":10123,"duration_ms":96459,"significance":"If the proofs are completed, the paper gives the first explicit constant terms for the total integrals of this one-parameter family of tronquée Painlevé II solutions and their associated Hamiltonians. The algebraic identities in Lemma 4.1 are clean and checkable, and the agreement with the known special cases, including the Tracy-Widom constant c0 and the previously known α=0, ω=1 Hamiltonian integral, provides strong independent evidence that the final formulas are correct. The paper also indicates plausible applications to large-gap asymptotics in random matrix theory. However, the proof of Theorem 1.3 currently has a load-bearing gap concerning parameter uniformity, and one displayed identity in the derivation contains a factor error that needs correction.","major_comments":[{"comment":"The proof of Theorem 1.3 differentiates the regularized integrals with respect to α or β and then integrates asymptotic relations over parameter intervals. Propositions 3.1-3.4 state asymptotic estimates for fixed parameters only; no uniformity in α on compact subsets of (-1/2,∞) or in β on compact subsets of iR is stated for the O(|s|^{-3/2}) remainders, and the derivatives ∂αw and ∂βw that enter (4.5)-(4.6) are not among the quantities estimated in those propositions. This is not a harmless technicality: equation (4.15) is obtained by integrating (4.14) from 0 to α, and equations (4.22) and (4.25) integrate in β and α respectively. Without a uniformity statement (or an alternative argument controlling the integrated remainders), the passage from (4.14) to (4.15) and from (4.24) to (4.25) is not justified. This step fixes the constant terms in (1.34)-(1.35), so it is load-bearing.","section":"§4.3, Eqs. (4.10)-(4.11), (4.14)-(4.15), (4.21)-(4.25)"},{"comment":"The first displayed identity for u(s)w_α(s) is incorrect as printed. Since (3.76) gives w=|s|^{1/2}tan B+O(s^{-1}) with B=θ/2+argΓ(1+α−β)−π/4, differentiation yields w_α=|s|^{1/2}sec^2B·dαB+...; combined with (3.75) this gives u w_α = 2|α−β| cos(A)/cos(B)·dαB + O(...), where A=θ/2+argΓ(α−β)+π/4. The printed formula has cos(A)cos(B) instead of cos(A)/cos(B). The following line, (2iβ−2α tanB)dαB, is consistent with the corrected quotient form when β is purely imaginary, so the error appears repairable, but the derivation of (4.25) must be corrected.","section":"§4.3, Eq. (4.23)"}],"minor_comments":[{"comment":"The condition written as 'βi ∈ R' is confusing and should be stated as iβ ∈ R or β ∈ iR.","section":"Theorems 1.1 and 1.3 and throughout"},{"comment":"The abstract contains a typo: 'Painelv\\'e' should be 'Painlev\\'e'.","section":"Abstract"},{"comment":"In the sentence beginning 'The asymptotics of the tronquée solution w(s;2α+1/2,ω) in (3.61)...', the reference should be to (3.76), since (3.61) is the separate ω=0 case.","section":"§3.4, proof of Proposition 3.4"},{"comment":"The paper switches freely between I1 and the rescaled integral ~I1; a short sentence reminding the reader of the exact relation (4.3) when (4.12)-(4.13) are used in the proof of Theorem 1.1 would improve readability.","section":"§4.2 and §4.3"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is my take on Dai–Xu–Zhang (arXiv:1908.01532). The headline: the main formulas are likely correct, and this is a genuinely useful addition to the Painlevé integral literature. The authors derive explicit constant terms for regularized integrals of tronquée PII solutions and their Hamiltonians for α > -1/2 and ω ≥ 0, with the constants expressed through Gamma, Barnes G, and ζ'(-1). This is the hard part in such expansions, and the paper unifies earlier special cases from Bothner–Its–Prokhorov, Baik–Buckingham, and Deift–Its–Krasovsky.\n\nWhat is done well: the differential identities in Lemma 4.1 are clean, and the paper pins the integration constants using external results rather than circularly. The special-case checks are meaningful independent evidence. The RH analysis itself is imported from earlier papers, but the integration over parameters is the actual new step.\n\nWhere I have reservations: the proof of Theorem 1.3 differentiates the asymptotic relations with respect to α and β and integrates over parameter intervals. The stated O(|s|^{-3/2}) error terms are only for fixed parameters; no uniform-in-parameter bound is proved. Differentiating the oscillatory asymptotics for w in (3.76) is not justified by the stated error alone. This is a genuine gap, but it looks fillable and the final formulas match known special cases, so the result is probably sound.\n\nA smaller issue: in (4.23) the first displayed expression has cos(A)cos(B), but differentiating w = sqrt(|s|) tan B gives cos(A)/cos(B). The next line writes the equivalent ratio form, so this is likely a typo, but it should be fixed.\n\nThe phase-transition discussion is explicitly heuristic and is fine.\n\nBottom line: this paper deserves a serious referee. The missing uniformity is the main thing to fix, technical rather than fatal. I would expect acceptance after a revision, and the explicit constants are worth citing.","headline":"Solid extension of Painlevé II integral asymptotics; the constant terms are new and match known cases, but the proof of Theorem 1.3 needs a uniformity argument and Eq. (4.23) has a misprinted factor.","tokens_in":38315,"tokens_out":3553,"would_cite":true,"duration_ms":32267,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["34M55","33E17","41A60","35Q15"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves explicit asymptotic constants for integrals of Painlevé II tronquée solutions and their Hamiltonians.","keywords":["Painlevé II equation","tronquée solutions","Hamiltonian","Riemann-Hilbert problem","asymptotic expansion","Barnes G-function","total integrals","random matrix theory"],"falsifier":"One concrete check: evaluate $I_2(s;1/4,0)$ by direct quadrature of (1.33) at a large negative $s$, say $s=-100$, and compare the result with (1.34); a mismatch in the constant term $\\ln(G(3/2)/(2\\pi)^{1/4})-\\zeta'(-1)+1/4-\\ln 2/24$ larger than the stated $O(|s|^{-3/2})$ would refute Theorem 1.3. A special-case check is $I_2(s;0,1)=0$, which the formula with $\\alpha=0$, $\\beta=0$ predicts exactly.","tokens_in":37215,"feed_emoji":"🧮","tokens_out":9640,"duration_ms":87663,"temperature":0.7,"pith_summary":"This paper proves that two regularized integrals attached to the tronquée solutions of the Painlevé II equation, one for the solution $q$ and one for its Hamiltonian $H$, are convergent and independent of the cutoff $c$, for $\\alpha > -1/2$ and $\\omega \\ge 0$. It then computes their full asymptotic expansions as $s$ tends to $+\\infty$ and $-\\infty$, with all constant terms expressed through the Gamma function, the Barnes $G$-function, $\\zeta'(-1)$ and $\\arg \\Gamma$. The formulas specialize to the known total-integral results for the Hastings-McLeod and Ablowitz-Segur solutions, and the Hamiltonian integral recovers the constant in the large-gap expansion for a Gaussian unitary ensemble with an edge discontinuity. A sympathetic reader should care because explicit constants in such integrals are the difficult part of many random-matrix large-gap asymptotics, and the paper supplies them for a whole one-parameter family of Painlevé II solutions at once.","feed_headline":"Tronquée Painlevé II integrals now have exact constants","feed_subtitle":"Two regularized integrals, for the solution and its Hamiltonian, get full asymptotics with Gamma and Barnes G constants.","key_machinery":"The argument rides on a Riemann-Hilbert (RH) problem for the Painlevé XXXIV equation, whose solution $\\Psi$ encodes the Hamiltonian $H$ and the tronquée solution $w$ through $\\ln|(\\Phi_0)_{11}|$. Large-$s$ asymptotics of $w$, $H$, $u$, and $\\ln|(\\Phi_0)_{11}|$ are obtained by nonlinear steepest descent with Airy, Bessel, and confluent hypergeometric parametrices. The proof then uses differential identities (2.19)-(2.21) to express derivatives of an auxiliary integral $I_3$ with respect to $\\alpha$ and $\\beta$ in terms of $I_1$, $u$, and $w$, and the bridge identities of Lemma 4.1, especially $I_2 = -\\frac{1}{3}(uw+2sH+2\\alpha^2+\\alpha)+2\\alpha I_1-I_3$, to pin down the integration constants.","core_discovery":"The central claim is that the regularized integrals $I_1(s;\\alpha,\\omega)$ of the tronquée solution and $I_2(s;\\alpha,\\omega)$ of the associated Hamiltonian are well defined and have explicit two-ended asymptotics. For $\\omega=0$ the constants involve $\\ln \\Gamma(1+2\\alpha)$ and $\\ln(G(1+2\\alpha)/(2\\pi)^\\alpha)$ respectively; for $\\omega=e^{-2\\beta\\pi i}$ with $\\beta i\\in\\mathbb{R}$, they involve $\\arg\\Gamma$, a cosine of the phase $\\vartheta(s)$, and the combination $G(1+\\alpha+\\beta)G(1+\\alpha-\\beta)/G(1+2\\alpha)$. Theorem 1.1 states expansions (1.20)-(1.21), and Theorem 1.3 states (1.34)-(1.35). The paper also shows the special cases agree with earlier results: $\\alpha=-1/4$, $\\omega=0$ recovers the Hastings-McLeod integral, and $\\alpha=0$ recovers the large-gap constant in the perturbed-GUE setting.","pith_inferences":["Editorial inference: the integration-over-parameters step that fixes the constants assumes the $O(|s|^{-3/2})$ error terms in Propositions 3.1-3.4 are uniform in $\\alpha$ and $\\beta$; the paper states the asymptotics pointwise in the parameters and would need a uniformity argument to fully justify (4.25).","Editorial inference: Remark 3.2 indicates the RH analysis extends to $\\omega \\in \\mathbb{C}\\setminus(-\\infty,0)$, so the integral formulas probably extend to complex $\\beta$ away from the negative axis, with the absolute values interpreted analytically.","Editorial inference: the Hamiltonian integral $I_2$ is closely tied to the logarithm of an isomonodromic tau function, so the Barnes-function constants found here are likely the explicit connection formulas for tau functions in this Painlevé II family.","Editorial inference: a direct numerical quadrature of (1.33) at $\\alpha=1/4$, $\\omega=0$ would test the constant term independently of the RH machinery."],"forward_implications":["Setting $\\omega=0$ in (1.20) and $\\alpha=-1/4$ reproduces the known total integral of the Hastings-McLeod solution, so the new formulas contain that classical result.","Setting $\\alpha=0$ in (1.35) gives another proof of the large-gap expansion for the Gaussian weight with an edge discontinuity, including the constant $\\ln G(1+\\beta)G(1-\\beta)-3\\beta^2\\ln 2$.","The Hamiltonian integral supplies the explicit constant $\\ln(G(1+2\\alpha)/(2\\pi)^\\alpha)$ that turns the paper's soft-to-hard edge phase-transition heuristic into a concrete statement about the transition of the gap probability."],"supporting_citations":[{"why":"supplies the nonlinear steepest-descent analysis of the Painlevé XXXIV RH problem for large positive and negative $s$ that Propositions 3.1 and 3.4 rely on.","marker":"[33]"},{"why":"provides existence and pole-freeness of the RH solution, the relation $H=s^2/4+i(\\Psi_1)_{12}$, and the large-$s$ asymptotics of $u$ and $H$ for $\\omega=0$ and $\\omega>0$.","marker":"[51]"},{"why":"proves the Tracy-Widom large-gap constant $c_0=\\frac{1}{24}\\ln 2+\\zeta'(-1)$ used to fix $I_3(s;0,0)$ in (4.17).","marker":"[15]"},{"why":"establishes the total integral formulas for Ablowitz-Segur and Hastings-McLeod solutions that Theorem 1.1 must reproduce at special parameter values.","marker":"[4]"},{"why":"gives the comparison formula (1.36) for $\\alpha=0$, $\\beta i\\in[0,+\\infty)$, used both as a consistency check and as the source of the Barnes-function combination.","marker":"[9]"},{"why":"conjectures the large-gap expansion in (1.42), whose proof via (1.35) is one of the paper's stated applications.","marker":"[5]"}],"fun_headline_variants":["Exact constants for Painlevé II tronquée integrals and Hamiltonians","Exact asymptotic constants for tronquée Painlevé II integrals","Tronquée Painlevé II: exact integrals and Hamiltonian constants","Exact Gamma and Barnes G constants for Painlevé II integrals","Painlevé II tronquée integrals: exact constants via Gamma and G"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof differentiates the regularized integrals with respect to $\\alpha$ and $\\beta$ and integrates the resulting asymptotic relations over parameter intervals, but it does not explicitly prove that the error terms are uniform in those parameters on the ranges used.","fun_headline_variants_meta":{"raw":{"variants":["Exact constants for Painlevé II tronquée integrals and Hamiltonians","Exact asymptotic constants for tronquée Painlevé II integrals","Tronquée Painlevé II: exact integrals and Hamiltonian constants","Exact Gamma and Barnes G constants for Painlevé II integrals","Painlevé II tronquée integrals: exact constants via Gamma and G"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000848,"raw_usage":{"total_tokens":3712,"prompt_tokens":989,"completion_tokens":2723,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":2624}},"tokens_in":605,"tokens_out":2723,"duration_ms":18350,"temperature":1.0,"reasoning_tokens":2624,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:10:43.973796+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One concrete check: evaluate $I_2(s;1/4,0)$ by direct quadrature of (1.33) at a large negative $s$, say $s=-100$, and compare the result with (1.34); a mismatch in the constant term $\\ln(G(3/2)/(2\\pi)^{1/4})-\\zeta'(-1)+1/4-\\ln 2/24$ larger than the stated $O(|s|^{-3/2})$ would refute Theorem 1.3. A special-case check is $I_2(s;0,1)=0$, which the formula with $\\alpha=0$, $\\beta=0$ predicts exactly.","supporting_citations":[{"cited_title":"Its, A.B.J","cited_arxiv_id":null,"evidence_quote":"supplies the nonlinear steepest-descent analysis of the Painlevé XXXIV RH problem for large positive and negative $s$ that Propositions 3.1 and 3.4 rely on."},{"cited_title":"Wu, S.-X","cited_arxiv_id":null,"evidence_quote":"provides existence and pole-freeness of the RH solution, the relation $H=s^2/4+i(\\Psi_1)_{12}$, and the large-$s$ asymptotics of $u$ and $H$ for $\\omega=0$ and $\\omega>0$."},{"cited_title":"Deift, A","cited_arxiv_id":null,"evidence_quote":"proves the Tracy-Widom large-gap constant $c_0=\\frac{1}{24}\\ln 2+\\zeta'(-1)$ used to fix $I_3(s;0,0)$ in (4.17)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"establishes the total integral formulas for Ablowitz-Segur and Hastings-McLeod solutions that Theorem 1.1 must reproduce at special parameter values."},{"cited_title":"Bothner, A","cited_arxiv_id":null,"evidence_quote":"gives the comparison formula (1.36) for $\\alpha=0$, $\\beta i\\in[0,+\\infty)$, used both as a consistency check and as the source of the Barnes-function combination."},{"cited_title":"Bogatskiy, T","cited_arxiv_id":null,"evidence_quote":"conjectures the large-gap expansion in (1.42), whose proof via (1.35) is one of the paper's stated applications."}],"review_version":1}