{"id":"eefffdd1-c290-403c-9b46-f257fc138b35","arxiv_id":"2507.07282","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A deformed RSJ model related to general Heun equations keeps integer-only phase-lock areas while breaking all constrictions.","lead":"Two new families of dynamical systems on a torus are introduced that correspond to general and confluent Heun equations, generalizing the resistive-shunted Josephson junction (RSJ) model. In the main deformed family, rotation-number quantization still holds, but the characteristic constriction points of the phase-lock areas disappear.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Constriction-breaking conclusion rests on unproved extension of [20, Prop. 2.2] to all monotone Möbius families; Theorem 24 only excludes identity monodromy, not parabolic nonidentity self-intersections.","rationale":"This is the single most load-bearing concern because every step of the paper leading to the advertised new phenomenon, constriction breaking, passes through Corollary 25. The Heun correspondence and the residue-eigenvalue computation in Theorem 24 are carefully done and appear correct for δ∈(0,1); the rotation-number quantization argument is standard. The missing piece is not an internal inconsistency but an unproved transfer of a special lemma. The reader identified exactly this assumption, and the requested proof or a counterexample is a reasonable condition for acceptance. I do not see a reason to move the verdict to accept or reject; keeping the conditional verdict is appropriate. The numerical check proposed would give direct evidence about whether dRSJ actually develops parabolic nonidentity self-intersections, and the re-derivation would settle the lemma's validity.","tokens_in":22615,"tokens_out":15304,"duration_ms":179899,"concrete_test":"Independently re-derive [20, Prop. 2.2] for an arbitrary strictly increasing analytic family of Möbius circle diffeomorphisms, and check whether a constriction can occur at a parabolic nonidentity map. In parallel, perform a numerical search in the dRSJ family at δ=0.5, ω=1, D=0: for the r=1 phase-lock area, compute the boundary branches B_-(A), B_+(A) by high-precision integration and locate any A≠0 where they meet; at each meeting point, compute the 2×2 monodromy of system (2.4). A non-scalar monodromy with trace ±2 would be a parabolic nonidentity constriction and would refute Corollary 25; scalar monodromy at every meeting point supports the claimed mechanism.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline qualitative claim is that in dRSJ (2.1) with δ≠0 all constrictions break. Corollary 25 deduces this from Theorem 24 via the assertion that [20, Prop. 2.2] 'remains valid in full generality together with its proof.' That proposition, proved only for the RSJ family, says every constriction of a phase-lock area corresponds to a trivial (identity) Poincaré map. This step is load-bearing because Theorem 24(1) establishes only the absence of identity monodromy for A≠0, δ∈(0,1). In a general strictly increasing family of Möbius circle maps, the boundary of a phase-lock area is the parabolicity locus (trace ±2); a self-intersection of that locus can in principle occur at a parabolic nonidentity map with a single double fixed point. Such a point would be a constriction not excluded by Theorem 24. The RSJ identification of constrictions with identity uses special structure (e.g., constriction alignment B=rω), so the blanket extension is not self-evident. The 'Thus' clause also assumes phase-lock areas retain the RSJ garland topology in which constrictions are the only possible separators of components in half-planes; this is not proved. There is also a minor statement mismatch: Theorem 24(1) says δ∈[0,1), while the proof and the known RSJ behavior require δ∈(0,1).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs two families of torus dynamical systems associated, respectively, with general and confluent Heun equations, obtained by projectivizing linear systems of a special 'torus dynamical type.' The main object is the deformed RSJ family (2.1), dθ/dτ = (cosθ + B + A sinτ)/(ω(1 − δ cosτ)) + D. The authors prove that this family is equivalent to a family of general Heun equations (Theorem 17 and Proposition 22) and that a confluent analogue exists (Theorem 20 and Proposition 23). They also prove rotation-number quantization for (2.1) (Theorem 2) and a monodromy criterion for the identity Poincaré map (Theorem 24). Corollary 25 then asserts that, for δ ≠ 0, all constrictions break and the phase-lock areas become connected in the upper and lower half-planes. The paper also contains a computation of growth points (Theorem 26) and a rotation-number formula for the restricted family (Proposition 27).","tokens_in":22933,"tokens_out":8660,"duration_ms":95744,"significance":"If the constriction-breaking claim is fully established, the paper identifies a clean qualitative difference between the RSJ model and its general-Heun deformation, and it connects Heun-equation monodromy to phase-lock topology in a way that goes beyond the previously studied double-confluent case. The paper contains substantial direct algebraic work: the residue eigenvalue computation in Theorem 24, the reduction of linear systems to Heun equations, and the explicit rotation-number formula are concrete and mostly verifiable by direct calculation. The numerical figures provide a falsifiable prediction (absence of constrictions for δ ≠ 0). The main monodromy theorem is not circular and appears sound. However, the headline geometric conclusion is conditional on an unproved structural lemma, so the significance of the paper is not yet fully realized.","major_comments":[{"comment":"The central claim that all constrictions break depends entirely on the assertion that [20, Proposition 2.2] 'remains valid in full generality together with its proof.' This is a load-bearing unproved premise. Theorem 24(1) excludes only identity Poincaré maps, whereas a constriction of a phase-lock area in a general strictly B-monotone family of Möbius circle maps could, in principle, occur at a nonidentity parabolic map with a single double fixed point. The RSJ proof of [20, Proposition 2.2] uses special structure (for example, constriction alignment B = rω), so the generalization is not self-evident. Please either supply a proof of the generalized constriction lemma or weaken Corollary 25 accordingly.","section":"§5.3, Corollary 25"},{"comment":"The connectedness conclusion of Corollary 25 also assumes, without proof, that the phase-lock areas of the deformed family (2.1) retain the RSJ 'garland' topology, in which constrictions are the only possible separators of components in the upper and lower half-planes. Even if one granted the absence of constrictions, the interiors of phase-lock areas could in principle remain disconnected for other reasons. A separate argument, at least for sufficiently small δ, is needed to justify the transition from 'no constrictions' to 'connected intersections with half-planes.'","section":"§5.3, Corollary 25"}],"minor_comments":[{"comment":"The wording 'Let δ ∈ [0, 1) and A ∈ R be both non-zero' is ambiguous and, if read literally as allowing δ = 0, contradicts the known existence of identity Poincaré maps at RSJ constrictions. The statement should say δ ∈ (0, 1) and A ≠ 0.","section":"§5.3, Theorem 24(1)"},{"comment":"Theorem 24(2) does not state a domain for δ, but the proof uses δ ∈ (0, 1). Please state the hypotheses explicitly, since the residue computation does not cover the δ = 0 case.","section":"§5.3, Theorem 24(2)"},{"comment":"References [7] and [8] are identical; one of them should be removed or replaced by the intended distinct citation.","section":"References"},{"comment":"The proof of the rotation-number formula (5.24) is sketched via an argument modulo Z and sign followed by monotonicity; it would be helpful to spell out the continuity and monotonicity argument in a few more lines.","section":"§5.3, Proposition 27"},{"comment":"The parameter δ is introduced as belonging to (0, 1) in (2.1), but later results and the RSJ limit use δ = 0; making the domain of δ consistent throughout would remove a source of confusion.","section":"§2.1, Eq. (2.1)"}],"recommendation":"major_revision","confidential_remarks":"The paper is suitable in scope for a dynamical systems journal, and the direct algebraic parts are valuable. The main reservation is that the headline result, constriction breaking, is conditional on an unproved generalization of the authors' earlier Proposition 2.2. If the authors can supply a proof of that generalized lemma, or reformulate the conclusions to match what Theorem 24 alone establishes, the paper would be considerably stronger. The statement mismatch in Theorem 24(1) should also be corrected before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this paper. The genuine novelty is the construction: two new families of torus dynamical systems (dRSJ and confluent) that are projectivizations of linear systems equivalent to general and confluent Heun equations. The classification theorems (17 and 20) are new, and the algebraic proofs (Propositions 22, 23, and Theorem 24's residue computation) check out; I verified the key eigenvalue relations (5.9)-(5.10) and the rotation number formula (5.24) matches the δ=0 limit. The quantization effect (Theorem 2) is a clean application of known Möbius results. This is a real bridge between Heun equations and phase-locking, and the dRSJ deformation is a sensible object to study.\n\nThe soft spot is the constriction-breaking claim. Corollary 25 leans on an unproved blanket extension of [20, Prop. 2.2] from the RSJ family to every strictly B-monotone family of Möbius circle maps. Theorem 24(1) only shows that for δ∈(0,1), A≠0, the Poincaré map is never the identity. But a constriction could in principle be a parabolic nonidentity map with a single double fixed point; nothing in the paper excludes that. The 'Thus' clause also assumes phase-lock areas still have the RSJ garland topology, which is not proved. So the headline 'constrictions break' is not fully established. It is plausible and probably true, but the proof as written has a gap.\n\nThere's also a minor mismatch: Theorem 24(1) includes δ=0, while the proof needs δ>0 and the RSJ limit has known constrictions, so the statement should be restricted to δ∈(0,1).\n\nNone of this is fatal to the core. The Heun classification and the monodromy analysis are solid and independently useful. The constriction-breaking is the main advertised punchline, though, and right now it is more a conjecture with strong supporting evidence than a theorem. The authors should either prove the extended lemma or soften the claim.\n\nWho should read it: people working on Josephson junction models, rotation number quantization, or Heun equations. I'd bring it to the reading group. I would not cite the constriction-breaking as a theorem until the gap is closed, but I'd cite the classification and quantization results. It deserves a serious referee: the ideas are new, the hard computations are mostly correct, and the gap is addressable. Send it to peer review with a request to fix the constriction lemma.","headline":"New families connecting Heun equations to torus dynamics are solid and worth publishing, but the headline constriction-breaking result rests on an unproved extension of a prior lemma.","tokens_in":23471,"tokens_out":5317,"would_cite":true,"duration_ms":48669,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["34M35","37E10","37E45"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that deforming the RSJ Josephson model to a general-Heun family preserves rotation number quantization but breaks every constriction, so phase-lock-area interiors become connected.","keywords":["Heun equation","phase-lock areas","rotation number quantization","constrictions","Josephson junction","torus dynamical systems","monodromy","Riccati equation"],"falsifier":"Compute the Poincaré map of (2.1) numerically for fixed $\\delta=0.5$, $\\omega=1$, $D=0$ with $A=0.5$ and scan $B$ over a fine grid; if any $B$ gives the identity map, Theorem 24 fails. Alternatively, for the same parameters plot the $r$-th phase-lock area and check whether its intersection with a horizontal line $A=0.5$ is disconnected by a point; a surviving constriction would disprove Corollary 25. A direct check of the $A=0$ prediction $B^2-1=\\omega^2(1-\\delta^2)(D-n)^2$ by integrating (5.14) also settles the identity-map locus.","tokens_in":22365,"feed_emoji":"⚡","tokens_out":6825,"duration_ms":69945,"temperature":0.7,"pith_summary":"The paper introduces the deformed RSJ family of torus dynamical systems and proves that two hallmarks of the original RSJ model separate: the rotation-number quantization effect survives, while the constrictions that pinch each phase-lock area into a garland disappear. Concretely, in the family $\\frac{d\\theta}{d\\tau} = \\frac{\\cos\\theta + B + A\\sin\\tau}{\\omega(1-\\delta\\cos\\tau)} + D$ with $\\delta \\in (0,1)$, phase-lock areas still exist only at integer rotation numbers, but for $A \\neq 0$ the Poincar\\'e first-return map is never the identity, so no constrictions can occur. The paper also builds a confluent-Heun companion family and identifies the precise curves on the $A=0$ axis where identity maps and growth points can appear. A sympathetic reader would care because this is the first qualitative separation between the RSJ model and a Heun-equation deformation of it, and it gives a monodromy-based mechanism for when phase-lock areas are connected versus chained.","feed_headline":"Deformed Josephson model keeps quantization but breaks constrictions","feed_subtitle":"A time-varying critical current removes the pinch points that chain Shapiro steps, while integer-step quantization survives","key_machinery":"The engine of the argument is the class of torus dynamical type linear systems: Fuchsian (or confluent) systems on the Riemann sphere whose projectivized Riccati equation preserves the product of unit circles $S^1_\\Phi \\times S^1_z$. Restricting to those circles via $\\Phi=e^{i\\theta}$, $z=e^{i\\tau}$ turns the Riccati equation into the torus ODE (2.1). Because the Poincar\\'e map of such an ODE is the projectivized monodromy of the linear system, it is a M\\\"obius transformation, and an identity Poincar\\'e map is equivalent to scalar monodromy. The proof of Theorem 24 compares residue eigenvalues at the singularities $0$ and $\\alpha$, obtaining the relation $(\\nu+n)^2=(\\bar{\\nu}+c)^2-4b^2$, which forces $\\operatorname{Im}\\nu=0$ and hence $A=0$; this is what rules out off-axis identity maps. The constriction-breaking conclusion then rides on the lemma that constrictions sit only at trivial Poincar\\'e maps.","core_discovery":"The central claim is that the deformed RSJ family (2.1) inherits the rotation number quantization of the RSJ model but loses its constrictions. Theorem 2 states that phase-lock areas exist only for integer rotation numbers, because the family is strictly increasing in $B$ and its Poincar\\'e maps are M\\\"obius transformations. Theorem 24 states that for $\\delta \\in (0,1)$ and $A \\neq 0$ the Poincar\\'e map is never the identity, and that for $A=0$ identity maps occur exactly along $B^2-1=\\omega^2(1-\\delta^2)(D-n)^2$ with $n \\in \\mathbb{Z}$. Since a constriction, by the lemma quoted from [20], can occur only at a trivial Poincar\\'e map, Corollary 25 concludes that in the dRSJ family all constrictions break and the intersections of phase-lock-area interiors with the upper and lower half-planes become connected. On the $A=D=0$ axis the paper also computes the remaining growth points and the rotation number $\\rho = \\frac{\\sqrt{B^2-1}}{\\omega\\sqrt{1-\\delta^2}}$.","pith_inferences":["The same monodromy-and-constriction logic likely applies to the confluent companion family (2.7), so its phase-lock areas may also have connected interiors; the paper does not state this.","The explicit identity-condition curves give a testable numerical signature: for fixed $\\omega$, $\\delta$, and $D$, scanning $B$ at $A=0$ should show degenerate or growth points exactly at $B^2-1=\\omega^2(1-\\delta^2)(D-n)^2$, and deviation would indicate that the extended constriction lemma fails.","If a physical Josephson device realizes the dRSJ denominator as a time-modulated critical current, Shapiro steps should appear with connected step interiors rather than the familiar chained garlands, a distinction visible in current-voltage curves.","Because constrictions correspond to collisions of monodromy data in the RSJ case, their breaking suggests that the isomonodromic foliation of dRSJ parameters has a different topology, possibly traceable through Painlev\\'e-type equations."],"forward_implications":["For every nonzero $A$, the Poincar\\'e map of the deformed RSJ family is not the identity when $\\delta \\in (0,1)$, so the family has no constrictions.","The interiors of each phase-lock area in the upper and lower half-planes become connected in the dRSJ family, unlike the garland chains of the RSJ model.","Rotation number quantization survives, so phase-lock areas in the deformed family still exist only for integer rotation numbers.","On the axis $A=0$, identity Poincar\\'e maps occur exactly along $B^2-1=\\omega^2(1-\\delta^2)(D-n)^2$, locating the growth points of the phase-lock areas in the $(B,D)$ plane.","For $A=D=0$ and $B>1$, the rotation number is explicitly $\\rho = \\frac{\\sqrt{B^2-1}}{\\omega\\sqrt{1-\\delta^2}}$."],"supporting_citations":[{"why":"Supplies the rotation-number quantization theorem for Möbius families strictly increasing in a parameter, which Theorem 2 invokes directly.","marker":"[11]"},{"why":"Provides Proposition 2.2, the constriction-trivial-Poincaré-map lemma that Corollary 25 extends to the dRSJ family.","marker":"[20]"},{"why":"Establishes the DCHE-monodromy description of phase-lock areas that motivates the GHE and CHE construction.","marker":"[6]"},{"why":"Relates the RSJ model to double confluent Heun equations and entire solutions, supporting the linear-system equivalence.","marker":"[5]"},{"why":"Gives the known constriction-alignment result $B=\\omega r$ for the RSJ model, against which the constriction-breaking result is measured.","marker":"[4]"}],"fun_headline_variants":["Deformed Josephson model: quantization persists, constrictions break","Josephson variant keeps integer phase-locks, drops constrictions","Torus flows from Heun equations lose constrictions, keep quantization","dRSJ model: phase-lock areas still integer, sharp points gone","Constriction breaking in generalized Josephson dynamics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a constriction of a phase-lock area can occur only where the Poincaré first-return map is the identity; the paper extends this lemma from the RSJ case to the deformed family by assertion rather than by a written proof.","fun_headline_variants_meta":{"raw":{"variants":["Deformed Josephson model: quantization persists, constrictions break","Josephson variant keeps integer phase-locks, drops constrictions","Torus flows from Heun equations lose constrictions, keep quantization","dRSJ model: phase-lock areas still integer, sharp points gone","Constriction breaking in generalized Josephson dynamics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000651,"raw_usage":{"total_tokens":3039,"prompt_tokens":1053,"completion_tokens":1986,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":669,"completion_tokens_details":{"reasoning_tokens":1900}},"tokens_in":669,"tokens_out":1986,"duration_ms":16884,"temperature":1.0,"reasoning_tokens":1900,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:45:29.504945+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Poincaré map of (2.1) numerically for fixed $\\delta=0.5$, $\\omega=1$, $D=0$ with $A=0.5$ and scan $B$ over a fine grid; if any $B$ gives the identity map, Theorem 24 fails. Alternatively, for the same parameters plot the $r$-th phase-lock area and check whether its intersection with a horizontal line $A=0.5$ is disconnected by a point; a surviving constriction would disprove Corollary 25. A direct check of the $A=0$ prediction $B^2-1=\\omega^2(1-\\delta^2)(D-n)^2$ by integrating (5.14) also settles the identity-map locus.","supporting_citations":[{"cited_title":"Buchstaber, O","cited_arxiv_id":null,"evidence_quote":"Supplies the rotation-number quantization theorem for Möbius families strictly increasing in a parameter, which Theorem 2 invokes directly."},{"cited_title":"Glutsyuk, V","cited_arxiv_id":null,"evidence_quote":"Provides Proposition 2.2, the constriction-trivial-Poincaré-map lemma that Corollary 25 extends to the dRSJ family."},{"cited_title":"Buchstaber and A","cited_arxiv_id":null,"evidence_quote":"Establishes the DCHE-monodromy description of phase-lock areas that motivates the GHE and CHE construction."},{"cited_title":"Buchstaber and A","cited_arxiv_id":null,"evidence_quote":"Relates the RSJ model to double confluent Heun equations and entire solutions, supporting the linear-system equivalence."},{"cited_title":"Bibilo and A","cited_arxiv_id":null,"evidence_quote":"Gives the known constriction-alignment result $B=\\omega r$ for the RSJ model, against which the constriction-breaking result is measured."}],"review_version":1}