{"id":"3ad9e39d-96ec-44ee-b25a-c2df87446fde","arxiv_id":"2505.09240","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Exact WKB is used to compute geometric and higher-order exponential factors for twisted and nonlinear Landau-Zener transitions, including a predicted phantom transition without a level crossing.","lead":"This paper applies the exact WKB method, a rigorous form of semiclassical approximation, to compute transition exponents in generalized and nonlinear Landau-Zener models. It derives a new higher-order 'quasi-geometric' exponent and argues that the standard linear approximation at level crossings can be unreliable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The all-order exactness claim rests on an unproved vanishing lemma for Sodd,n (n≥2); the displayed recurrence (23) governs the full Riccati coefficient, not the odd integrand, so the stated inductive regularity is not established.","rationale":"The reader's verdict is CONDITIONAL, and my independent read lands on the same load-bearing point. The paper is not claiming a mere asymptotic approximation; it claims that the three computed exponents are the complete transition exponents, with all Sodd,n (n≥2) contributions exactly zero. The only support is an appeal to an inductive regularity argument that is not written down. I checked the displayed equations around it: Eq. (23) governs the full Riccati coefficient S_n, whereas the contour integrals (44)-(48) use the odd combination Sodd,n. The recurrence for Sodd is nonlinear and second-order; the paper neither states it nor uses it, so the 'inductively' in footnote 4 cannot be verified from the manuscript. This is not an external-consensus disagreement; it is an internal gap in the justification of the central claim. The explicit residue computations for S_{-1}, S0, S1 (e.g., eqs. 78-80, 86-91) are detailed and plausible, and known limits (linear Landau-Zener) are reproduced, which is independent support. But that support does not cover n≥2. If a CAS computation of Sodd,2 for the double-twist model gives a nonvanishing residue, the all-order exactness claim fails and the transition matrix (58) is not exact; if the residue vanishes, the manuscript still needs to supply the induction, ideally as a lemma with the Sodd recurrence. Thus the conditional verdict is appropriate and no verdict change is needed.","tokens_in":19180,"tokens_out":23604,"duration_ms":243390,"concrete_test":"Symbolically expand the odd integrand F=Sodd for the double-twist model to order η^{-3} using F² - F''/(2F)+3(F')²/(4F²) = η²Q with Q0=-A²τ²-Λ², Q1=-3Aλτ³-iA, Q2=-(9/4)λ²τ⁴-3iλτ, and compute Res_{τ=∞}Sodd,2 (the coefficient of 1/τ in the Laurent expansion at infinity). If nonzero, the vanishing lemma in footnote 4 and the three-term exactness claim are false; if zero, repeat for θ=λτ⁴ to test whether the vanishing is a general theorem, and in either case supply the missing induction so the claim is checkable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that in the generalized and twisted Landau-Zener models the transition exponent is exactly the sum of the dynamical, geometric, and quasi-geometric terms, because every Sodd,n with n≥2 has zero contour integral around the MTP contour. The only support offered is footnote 4 and the remark before Eq. (58): 'It can be proven inductively... regular at x=∞,' with no proof. This is load-bearing: if the residue at infinity of, say, Sodd,2 is nonzero, the quasi-geometric term is not the last correction and Eq. (58) is not the exact transition matrix.\n\nThe concern is sharper than 'the proof is missing.' The recurrence written in Eq. (23) is for the coefficients of the full Riccati solution S in Eq. (17), i.e. for S^{(±)}, not for the odd combination Sodd=(S^{(+)}-S^{(-)})/2 that enters the contour integrals (44)-(48). The odd integrand satisfies a different nonlinear second-order equation, F²-F''/(2F)+3(F')²/(4F²)=η²Q, and its coefficients are not obtained by the linear-looking induction in Eq. (23). Thus the sketched induction cannot be checked from the equations the paper displays.\n\nThe regularity is also not trivially true. For the double-twist model φ=λτ³, Q0=-A²τ²-Λ², Q1=-3Aλτ³-iA, Q2=-(9/4)λ²τ⁴-3iλτ; Sodd,0 and Sodd,1 grow like τ² and τ³ at infinity, and S1 has a nonzero residue (eq. 88). Whether the next coefficient has a nonzero 1/τ term is exactly the unproved all-order assertion. A direct computation would settle it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies exact WKB analysis to a class of two-level Landau-Zener-type models and claims that the complete nonadiabatic transition exponent is exactly given by three residue contributions: the dynamical exponent from S_{-1}, the geometric exponent from S_0, and a new 'quasi-geometric' exponent from S_1, with all higher WKB coefficients contributing zero. The models treated include the original Landau-Zener transition, the twisted model φ=Bτ² and its shifted version, the double-twist model φ=λτ³, a model with non-linear diagonal elements, and a model with Δ=Λe^{iΘ(τ)/ℏ}. The paper also discusses the 'phantom' Landau-Zener transition and argues that conventional linear approximations at level crossings are unreliable. The central technical claim is that for these models higher-order odd WKB coefficients are regular at complex infinity, so their contour integrals vanish and the transition matrix retains the same form as the standard Landau-Zener matrix.","tokens_in":19609,"tokens_out":7369,"duration_ms":77461,"significance":"If the all-order vanishing lemma were proven, the paper would provide a valuable systematic exact-WKB derivation of Berry's geometric amplitude factors and an explicit second-order 'quasi-geometric' correction, going beyond Berry's first-order result. The residue-based method is computationally concrete and the explicit formulas for S_{odd,0}, S_{odd,1}, and the double-twist model are useful. The paper also makes a physically interesting observation about phantom Landau-Zener transitions and the unreliability of naive linear approximations. However, the advertised exactness for all orders rests on an assertion that is not proven in the manuscript, and several displayed central equations contain apparent algebraic or typographical errors. The contribution is therefore not yet in a form where its central claim can be accepted without additional work.","major_comments":[{"comment":"The exactness claim 'correct for all orders' depends entirely on the assertion that S_{odd,n} for n≥2 has zero residue at complex infinity for the models considered. This is stated in footnote 4 and again before Eq. (58) as something that 'can be proven inductively,' but no proof is given. Moreover, the recurrence displayed in Eq. (23) is the recurrence for the coefficients of the full Riccati solution S^(±), not for the odd combination S_{odd} defined in Eq. (24); the odd coefficient satisfies a different nonlinear equation, as follows from Eqs. (26)-(27). Therefore the stated induction cannot be checked from the equations in the paper. If, for example, S_{odd,2} had a nonzero 1/x coefficient, Eq. (58) would not be the exact transition matrix and the quasi-geometric term would not be the last correction. I ask the authors either to supply a proof of the vanishing lemma, ideally with explicit residue computations for the first few n≥2, or to weaken the all-order claim and state the result as valid up to the quasi-geometric order.","section":"Section III, after Eq. (48), footnote 4, and before Eq. (58)"},{"comment":"The integrands displayed in the geometric-exponent calculations do not match the Q1 obtained from the general formulas. For the twisted model φ=Bτ², using Eqs. (65)-(67) with Δ=Λe^{-iφ(τ)} and D=Aτ gives Q1 = -iA - 2ABτ², whereas Eq. (79) displays the numerator -A + 2Bτ² i. For the shifted model, Eq. (85) displays a denominator A²(τ²-τ0)², whereas the correct Q0 from the model has A²(τ-τ0)². Since these expressions determine the geometric exponents that are central results of the paper, the sign and factor conventions must be corrected and the final values recomputed. At present the displayed equations cannot reproduce the quoted results as written.","section":"Eqs. (79) and (85)"},{"comment":"The claim that the transition matrix keeps the same form (58) even when Q1(τ) and Q2(τ) are nontrivial functions is asserted rather than demonstrated. This is a load-bearing point because it justifies extracting the complete transition exponent from the contour integral of S_{odd} alone. The paper invokes the Borel-resummation starting point and the resulting connection formula, but no precise statement or reference is given for the generalized connection formula. A citation to the relevant exact-WKB theorem, or a short derivation in the present normalization, is needed before the 'same form' claim can be accepted.","section":"Section III A and Eq. (58)"},{"comment":"The double-twist model is presented as a case where the all-order exact WKB calculation is essential, but the paper provides no numerical or independent check of the quasi-geometric exponent, and the vanishing of all higher terms is again only asserted. Since this model is the main evidence for the new 'quasi-geometric' contribution, I recommend adding a direct check, for example by computing the residue of S_{odd,2} for this model or by comparing the exponent with a high-order numerical solution of the original two-level system.","section":"Section III C and Eq. (88)"}],"minor_comments":[{"comment":"The relation between the contour integral in Eq. (56), the coefficient -πκ, and the definition Γ_d ≡ -2πκ should be spelled out; as written, the factor 2 appears only in the definition and could confuse readers checking Eq. (56) against Eq. (55).","section":"Eq. (57)"},{"comment":"In the expansion of F(x), the penultimate line contains a term C4 = 16(-A²B²Λ² + iA⁴B - iA⁴B) in which the two imaginary terms cancel; this looks like a typographical error and should be corrected or simplified.","section":"Eq. (72)"},{"comment":"The expansion uses 'C1x' where the variable is τ; this is a minor notational inconsistency.","section":"Eq. (80)"},{"comment":"Several equations in Sections III B and III C have sign or notational inconsistencies (e.g., the numerator in Eq. (79), the denominator in Eq. (85), and the sign of the imaginary part in the displayed results). Even if the final real exponents are unaffected, the intermediate expressions should be made internally consistent.","section":"Throughout"},{"comment":"The abstract is much more general than the technical content of the paper and contains statements about 'universal phenomena' that are not directly established by the calculations; I suggest making the abstract more closely reflect the specific models and results.","section":"Abstract and Introduction"}],"recommendation":"major_revision","confidential_remarks":"The paper's central advertised result is the all-order exactness of the transition exponent. The main obstacle is not the framework, which is legitimate exact WKB, but the absence of a proof of the vanishing lemma for higher S_{odd,n}. If the authors can supply such a proof or reduce the claim to a verifiable computation, the paper would be publishable. Otherwise, the claim must be weakened and the remaining typos corrected. I would not recommend rejection at this stage, because the residue computations for the lower-order terms are explicit and largely correct in spirit, and the gap appears fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Useful read. The genuinely new content is the quasi-geometric S1 term for twisted and double-twisted Landau-Zener models and the phantom-transition analysis for nonlinear diagonals. The leading exponents reproduce Berry, so that is not the contribution; the paper earns its keep by giving an explicit residue-based workflow for higher corrections and by showing that the naive linear approximation at level crossings is unreliable. The Fig. 5 comparison looks convincing.\n\nWhat is good: the formulas for Sodd,-1, Sodd,0, Sodd,1 are explicit and checkable; the double-twist result (88) is a concrete new prediction; and the Stokes-line argument for \"phantom\" transitions is a nice conceptual point. The citation pattern is reasonable, including self-citations to the exact WKB papers where the framework is actually developed.\n\nThe main soft spot is the all-order claim. The vanishing of all higher Sodd,n is asserted but not supplied, and the stress-test point is exact: Eq. (23) is a recurrence for the full Riccati coefficients S_n, not for the odd combination Sodd,n that enters the contour integrals. The odd integrand satisfies a different nonlinear equation, so the displayed induction cannot be verified from the paper as written. This matters: if any Sodd,n with n≥2 has a nonzero residue at infinity, the quasi-geometric term is not the last correction and Eq. (58) is not exact. A direct computation of Sodd,2 for the double-twist model would settle it. Until then, the all-order statement should be flagged as a conjecture.\n\nSmaller issues: there are apparent sign and factor inconsistencies in Eqs. (57), (79), and (85). They look like typos, but they cost the reader time and should be fixed. There is also no numerical check of the quasi-geometric exponent, which is a shame because a one-line residue computation for Sodd,2 would make the central claim much more credible. Section E is explicitly preliminary and reads as speculation; it should be marked as such rather than presented as a result.\n\nVerdict: this deserves a serious referee. The main calculation is plausible and partly reproduces known results, but the paper should not be accepted until the vanishing lemma is either proven or explicitly stated as unproven, and the typos cleaned up. I would not cite the quasi-geometric exponent in my own work before that is settled. Conditional accept, not reject.","headline":"A useful but not yet rigorous extension of Berry's geometric amplitude calculation; the new quasi-geometric term is interesting, but the all-order claim needs proof.","tokens_in":20077,"tokens_out":2809,"would_cite":false,"duration_ms":29569,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["34M60","81Q20","81Q70"],"pacs":["03.65.Vf","03.65.Ta"],"model":"deepseek-v4-flash","headline":"For generalized and twisted Landau-Zener models, exact WKB gives the complete non-perturbative transition exponent as the sum of a dynamical term, a geometric term, and a quasi-geometric term, with all higher-order corrections vanishing.","keywords":["exact WKB","Landau-Zener transition","geometric exponent","quasi-geometric exponent","Borel resummation","Stokes phenomenon","non-perturbative transition","phantom Landau-Zener transition"],"falsifier":"Compute the next WKB coefficient $S_2$ for the cubic double-twist model and evaluate $\\mathrm{Res}_{\\tau=\\infty} S_2$; if the residue is nonzero in any parameter range, the claimed vanishing of higher orders is false. A numerical integration of the next-order term around the merged-pair contour for $D(t)=-a^2+b^2t^2$ would also settle whether the three-term exponent formula is complete.","tokens_in":18968,"feed_emoji":"⚫️","tokens_out":11718,"duration_ms":109465,"temperature":0.7,"pith_summary":"The paper aims to show that the exact WKB method—a Borel-resummed treatment of the Planck-constant expansion that fixes the Stokes-line structure exactly—computes the non-perturbative transition exponent in a wide family of two-level Landau-Zener models. In the models treated here, the total exponent splits into a dynamical part, a geometric part, and a new higher-order “quasi-geometric” part, and the paper claims that no further WKB corrections appear. This matters because the conventional way of deriving geometric amplitude factors uses a first-order expansion whose turning points drift when the Planck constant enters the Hamiltonian, and the paper argues that this makes the conventional calculation ambiguous. The exact WKB removes that ambiguity and also reveals a previously missing second-order contribution to the geometric exponent.","feed_headline":"Exact WKB completes the Landau-Zener exponent formula","feed_subtitle":"Dynamical, geometric, and quasi-geometric terms are all that contribute; higher corrections vanish.","key_machinery":"The central object is the exact WKB solution pair $\\psi_\\pm = S_{\\mathrm{odd}}^{-1/2} \\exp(\\pm \\int^x S_{\\mathrm{odd}} \\, dx')$, where $S_{\\mathrm{odd}}$ satisfies a Riccati equation and is expanded as $S_{\\mathrm{odd}} = \\eta \\sqrt{Q_0} + S_0 + \\eta^{-1} S_1 + \\cdots$. The argument is carried by the residue at complex infinity: the exponent along the merged-pair contour is $-\\mathrm{Re}\\,\\int_{\\gamma_\\pm} S_{\\mathrm{odd}}\\,dx = -\\mathrm{Re}\\,\\pi i \\,\\mathrm{Res}_{x=\\infty} [ S_{\\mathrm{odd}} ]$, so each order contributes only through the residue of $S_n$. The geometric part comes from $S_0 = Q_1/(2\\sqrt{Q_0})$, and the quasi-geometric part from the $S_1$ coefficient; for the models at hand the higher residues vanish, which is what makes the three-term formula exact.","core_discovery":"The central claim is that, after converting the two-level system into a Schrödinger equation with potential $Q = Q_0 + \\eta^{-1} Q_1 + \\eta^{-2} Q_2$, the total Landau-Zener exponent is the contour integral of the exact WKB function $S_{\\mathrm{odd}}$ around the merged pair of turning points, and only the first three terms $S_{-1}$, $S_0$, and $S_1$ contribute. The term $S_{-1}$ gives the dynamical exponent, $S_0$ gives the geometric exponent, and $S_1$ gives the quasi-geometric exponent. The paper claims this is exact for all orders, because every higher coefficient $S_n$ with $n \\ge 2$ is regular at complex infinity and therefore its integral vanishes. Explicit residue calculations for quadratic, cubic, and shifted twists, and for nonlinear diagonal elements, are used to show the structure and to expose the failure of the conventional linear approximation at level crossings.","pith_inferences":["Inference: if regularity at infinity is the actual mechanism, the same three-term exactness should hold for any two-level model with a quadratic $Q_0$ and polynomial $Q_1, Q_2$ of bounded degree; this gives a concrete test family the paper does not enumerate.","Inference: the quasi-geometric exponent can be read as the first $\\hbar$-correction to the geometric amplitude, suggesting a link to quantum-geometric tensors beyond the usual adiabatic curvature in driven materials.","Inference: for the phantom transition, a driven two-level system with no actual crossing of the diagonal levels should still show non-adiabatic tunneling or pair production with the same exponent as a real crossing, which would be a sharp test of the Stokes-line picture."],"forward_implications":["If the paper's claim is correct, the transition probability for the twisted Landau-Zener model is fully determined by three residue integrals, with no unknown higher-order corrections.","The geometric amplitude factor acquires a genuine second-order piece: for a quadratic twist with rate $B$, the quasi-geometric exponent is $\\pi B^2 \\Lambda^2/(2A^3)$ rather than zero.","For the cubic double-twist model, the geometric exponent can vanish while the quasi-geometric exponent remains nonzero, so the two corrections probe different features of the trajectory.","Shifting the twist away from the level crossing changes both geometric and quasi-geometric exponents, so the common statement that geometric factors are independent of the twist's location holds only when the twist sits at the crossing.","For nonlinear diagonal elements $D(t)=\\pm a^2 + b^2 t^2$, the conventional linear expansion near a crossing is numerically unreliable, and a “phantom” Landau-Zener transition with the same transition matrix occurs even without a level crossing."],"supporting_citations":[{"why":"supplies the first-order geometric amplitude factor that this paper extends and whose calculational ambiguity is exposed.","marker":"[17]"},{"why":"introduces the Borel-resummed exact WKB framework used throughout, including the connection formula and its higher-order correction.","marker":"[2]"},{"why":"provides the exact WKB connection formula for particle production that the paper adapts to the Landau-Zener problem.","marker":"[13]"},{"why":"gives a previous exact WKB treatment of Landau-Zener transitions, the closest methodological predecessor for the exponent integrals.","marker":"[14]"},{"why":"defines the original Landau-Zener crossing whose transition matrix is shown to persist in the generalized models.","marker":"[34]"},{"why":"supplies the twisted field-theory and semimetal Hamiltonian that motivates the quantized-field Landau-Zener model.","marker":"[30]"},{"why":"demonstrates exact WKB analysis of pair production by time-dependent electric fields, supporting the strong-field pair-production discussion.","marker":"[40]"}],"fun_headline_variants":["Exact WKB: only three terms govern Landau-Zener exponents","Three terms complete Landau-Zener exponent via exact WKB","Exact WKB yields exact exponents for nonlinear Landau-Zener","Dynamical, geometric, quasi-geometric: exact WKB's three terms","Exact WKB shows Landau-Zener exponent is three-term exact"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The all-order exactness rests on the paper's assertion, offered without proof, that every higher-order WKB coefficient is smooth enough at infinity that its contribution to the contour integral is zero.","fun_headline_variants_meta":{"raw":{"variants":["Exact WKB: only three terms govern Landau-Zener exponents","Three terms complete Landau-Zener exponent via exact WKB","Exact WKB yields exact exponents for nonlinear Landau-Zener","Dynamical, geometric, quasi-geometric: exact WKB's three terms","Exact WKB shows Landau-Zener exponent is three-term exact"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000793,"raw_usage":{"total_tokens":3505,"prompt_tokens":967,"completion_tokens":2538,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":583,"completion_tokens_details":{"reasoning_tokens":2443}},"tokens_in":583,"tokens_out":2538,"duration_ms":17215,"temperature":1.0,"reasoning_tokens":2443,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:37:51.582931+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the next WKB coefficient $S_2$ for the cubic double-twist model and evaluate $\\mathrm{Res}_{\\tau=\\infty} S_2$; if the residue is nonzero in any parameter range, the claimed vanishing of higher orders is false. A numerical integration of the next-order term around the merged-pair contour for $D(t)=-a^2+b^2t^2$ would also settle whether the three-term exponent formula is complete.","supporting_citations":[{"cited_title":"Quantal phase factors accompanying adi- abatic changes,","cited_arxiv_id":null,"evidence_quote":"supplies the first-order geometric amplitude factor that this paper extends and whose calculational ambiguity is exposed."},{"cited_title":"Schr¨ odinger equation","cited_arxiv_id":null,"evidence_quote":"introduces the Borel-resummed exact WKB framework used throughout, including the connection formula and its higher-order correction."},{"cited_title":"Reconstructing WKB from topological recursion","cited_arxiv_id":null,"evidence_quote":"provides the exact WKB connection formula for particle production that the paper adapts to the Landau-Zener problem."},{"cited_title":"The exact WKB for cos- mological particle production,","cited_arxiv_id":null,"evidence_quote":"gives a previous exact WKB treatment of Landau-Zener transitions, the closest methodological predecessor for the exponent integrals."},{"cited_title":"Topological contribu- tion to the Bogoliubov coefficient for cosmological parti- cle production,","cited_arxiv_id":null,"evidence_quote":"defines the original Landau-Zener crossing whose transition matrix is shown to persist in the generalized models."},{"cited_title":"Observation of the Geometric Amplitude Factor in an Optical System","cited_arxiv_id":null,"evidence_quote":"supplies the twisted field-theory and semimetal Hamiltonian that motivates the quantized-field Landau-Zener model."},{"cited_title":"Out-of-equilibrium crit- icalities in graphene superlattices,","cited_arxiv_id":null,"evidence_quote":"demonstrates exact WKB analysis of pair production by time-dependent electric fields, supporting the strong-field pair-production discussion."}],"review_version":1}