{"id":"aef31413-3297-471e-a3ff-09d60d99d5d8","arxiv_id":"2607.20099","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every admissible itinerary of windings around the two relativistic centres is realized by a collision-free orbit, including chaotic bounded, scattering, and trapped trajectories.","lead":"The paper shows that a relativistic version of the classic two-centre problem—one particle moving around two fixed masses—contains chaotic motion: any prescribed sequence of windings around the two centres is realized by a real, collision-free orbit. The authors prove this using energy-minimisation methods, and also construct scattering and trapped trajectories with prescribed behaviour.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proof of Theorem 2.1 rests on Proposition 6.3's unproved uniform-time claim; without it the collision minimisers Γ1, Γ2 that define the itinerary are not established, so the coding theorem lacks a foundation.","rationale":"The central claim is the coding theorem: every bi-infinite k-admissible sequence is realised by a true solution. The proof route is: construct two crossing-detector curves Γ1,Γ2 (collision minimisers), then show periodic minimisers in the corresponding homotopy classes have itineraries matching the sequence, and pass to a limit. The first and most foundational object is Γ1,Γ2; without them no itinerary can even be defined. Proposition 6.3 is the only construction of these curves, and its final step rests on an unproved and poorly justified uniformity assertion. This is a more serious gap than Lemma 5.3, whose confinement claim has a plausible (if terse) intersection argument, and than the curvature computation in Remark 6.7, which only affects the upgrade from semi-conjugation to conjugation in the R=0 case, not the existence of itineraries. The reader's verdict of CONDITIONAL is appropriate: the claimed theorems may well be true and the gaps may be repairable, but the proof as written does not establish them. My concern does not move the verdict; it reinforces it.","tokens_in":29994,"tokens_out":18732,"duration_ms":162707,"concrete_test":"Verify the uniform-time claim in Prop 6.3 in the simplest nontrivial case: the one-centre potential V(x)=μ/|x|^2 (μ>0, no perturbation, c2 absent) at energy h>0. For fixed q with |q|>r and a sequence q_n→0, solve the geodesic equation for the Jacobi metric (h+μ/|x|^2)(dx^2+dy^2) (e.g., numerically via a shooting method or discretised geodesic solver) and measure σ_n := last time the minimiser from q to q_n crosses ∂B_r. Check whether sup_n σ_n < ∞. If the sup is infinite, the assertion is false and Prop 6.3 collapses; if finite, the gap remains but the mechanism is plausible. An analytical version: derive the radial geodesic equation and prove a uniform bound on the time to reach radius r for all endpoints q_n→0; if such a bound cannot be derived without extra assumptions (e.g., monotonicity of the radius), the proof as written is incomplete.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The coding theorem (Theorem 2.1, hence Theorem 1.3) is built on the curves Γ1, Γ2, defined in §6.2 as collision minimisers from the centres to ∂D. Their existence is the content of Proposition 6.3. The proof of Prop 6.3 has a load-bearing gap. After proving local uniform convergence of fixed-end minimisers γ_n (from a fixed q to q_n→c) on compact time intervals, the proof must show the limit γ satisfies lim_{s→∞} γ(s)=c. This requires the assertion, stated without proof in §6.1, that 'for any r small enough there exists S_r such that γ_n([S_r,T_n]) ⊆ B_r' — i.e., that after a uniform time all minimisers remain inside every small ball around the centre. The attempted justification ('This implies that there exists at least a point of ∂B_r at infinite distance from q') is not a valid argument: a sequence of times S_k→∞ with γ_{n_k}(S_k)∉B_r does not produce a boundary point at infinite distance, and uniform-time control is exactly what is missing. Without this assertion, the diagonal limit γ may fail to converge to c, so the collision minimiser may not exist. If Γ1, Γ2 do not exist (or do not have the required non-intersection and transversality properties), the itinerary map π in (21) is not well-defined and the central claim that every k-admissible sequence is realised collapses. This is therefore the most load-bearing unproved step in the paper.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the planar relativistic two-centre problem at fixed energy, with potential V_rel = μ1/|q-c1| + μ2/|q-c2| + R(q). The authors show that, up to a time reparametrisation, the relativistic dynamics is orbitally equivalent to a classical two-centre system with a critical strong-force potential (V_rel + h_rel)^2. They then use variational methods for the Maupertuis functional to construct collision-free solutions with prescribed symbolic itineraries, proving a coding theorem (every bi-infinite k-admissible sequence is realised), a scattering theorem (solutions with prescribed asymptotic directions and finite admissible itineraries), and a trapping theorem (solutions asymptotically free in the past and prescribed symbolic behaviour in the future). The main theorems are stated both for the reduced classical system (Theorems 2.1--2.3) and, via the reduction, for the original relativistic system (Theorems 1.3--1.5).","tokens_in":30399,"tokens_out":9672,"duration_ms":83005,"significance":"If rigorously established, the results would be a significant step: they provide symbolic dynamics and chaotic subsystems for a non-integrable relativistic two-centre problem, extending the variational approach to critical strong-force singularities and complementing earlier work on classical N-centre problems. The reduction of the relativistic problem to a classical strong-force problem is elegant, and the proposed compactness and confinement strategy is natural. The paper explicitly builds on the authors' previous work [5], and the new material in Sections 4--7 is intended to adapt that framework to the present setting. However, several load-bearing arguments are only sketched or asserted, so the current manuscript is not yet convincing as a complete proof.","major_comments":[{"comment":"The proof of existence of collision minimisers contains a central unproved claim: after the diagonal limit γ is constructed, the text states 'minimizers must enter an Euclidean ball of fixed small radius in a uniform time' and claims that otherwise 'there exists at least a point of ∂B_r at infinite distance from q'. This justification is logically invalid: sequences n_k and S_k → ∞ with γ_{n_k}(S_k) ∉ B_r do not produce a point of ∂B_r at infinite distance. The uniform-time control is exactly what is needed to conclude lim_{s→∞} γ(s) = c. Without it, γ may fail to be a collision minimiser in the sense of Definition 6.1. Since Γ1 and Γ2 in §6.2 are constructed from these collision minimisers, and the itinerary map π in (21) depends on them, this gap directly affects Theorem 2.1 and hence Theorem 1.3. The claim needs a real proof, for example via a last-exit argument combined with the abse","section":"§6.1, Proposition 6.3"},{"comment":"The proof of confinement of all periodic minimisers to the disk D bounded by the [α1α2] minimiser is incomplete. The text says: 'assume by contradiction that a minimiser in H(τ) has one transversal intersection with φ. Then, it must intersect φ at least twice...' This only treats the case of at least one intersection. The zero-intersection case is not handled: the paper does not prove that a minimiser in a non-trivial class different from [α1α2] must intersect ∂D at all. Without that, the conclusion that its support lies entirely in D does not follow. This is load-bearing because Lemma 5.3 is used to define the compact region D and to justify the symbolic framework of Section 6.","section":"§5.1, Lemma 5.3"},{"comment":"The passage from the compactness limit γ_R → γ_∞ to the identification of the asymptotic velocity is not rigorous. After equation (22), the proof says 'Take an interval (-n,n) large enough so that |γ_∞(t)|=1/ε for some t∈(0,n)' and then derives an estimate with ε and O(ε^{β-1}). The logic is unclear, and the estimate does not establish lim_{t→+∞} \\dot γ_∞(t) = √(2h) e^{iθ_+}. A correct argument would need to compare \\dot γ_R(t) with √(2h)e^{iθ_+} uniformly on intervals whose time shifts go to infinity as R→∞, or use a different method to transfer the asymptotic direction from the approximating sequence to the limit. Since this is the core of the scattering theorem, the gap is substantial.","section":"§7.1, Proof of Theorem 2.2"},{"comment":"The trapping theorem is proved by a sketch rather than a complete argument. In particular, the statement 'From Lemma 5.3, we know that (|γ_∞(t)|>K for all t<0, |γ_∞(t)|<K for all t>0)' is not a consequence of Lemma 5.3 as stated. The proof also does not justify that the approximating minimisers γ_m have uniformly bounded times before entering B_K, nor that the fixed endpoint q in the construction does not prevent the limit from being globally defined and asymptotically free in the past. More details are needed for the convergence and for the transfer of the positively infinite itinerary to γ_∞.","section":"§7.2, Proof of Theorem 2.3"},{"comment":"The proof of Theorem 2.2 considers minimisers γ_{q^-,q^+}^R with endpoints on tangent lines ℓ± and a prescribed finite itinerary. However, Theorem 4.1 was proved only for endpoints satisfying |q^+|=|q^-| (condition (14)). The text asserts that the fixed-end results apply uniformly for endpoints outside B_{R0}, but the required extension to unequal radii is not proved. Since the existence of the approximate minimisers is the starting point of the scattering argument, this is a genuine gap, although it may be repairable by repeating the proof of Theorem 4.1 with a minor modification.","section":"§7.1, construction of minimisers between tangent lines"}],"minor_comments":[{"comment":"There are numerous typos and grammatical slips, e.g. 'rotation rotation' in Section 2, 'indipendently' in Lemma 5.6, 'mimimiser' in the caption of Figure 6, and 'Moreover,xhas' in the statement of Theorem 2.2. These should be corrected.","section":"Throughout"},{"comment":"The set T is described as 'the union of two open cylinders', but as defined it is a union of two one-dimensional submanifolds of the energy shell; the word 'open' is potentially misleading. Also, the decomposition into components C_i^± is not made precise.","section":"§6.2"},{"comment":"The computation of the negative curvature for R=0 contains several notational confusions, including the reuse of R for both the perturbation and a newly defined function in the displayed formula, and a typo in the denominator of the third term of R(q). The authors should rewrite this remark carefully, since the conjugation claim in the unperturbed case rests on it.","section":"Remark 6.7"},{"comment":"The proof delegates a key approximation and compactness step to [5, Lemma 6.11] with 'the same argument as'. Since this lemma is central to passing from finite periodic minimisers to a limit realising a non-periodic bi-infinite sequence, the authors should either state the lemma explicitly or reproduce the proof in sufficient detail for the present setting.","section":"Proof of Theorem 2.1"},{"comment":"In the proof of assertion iii), the argument that v_∞ must be parallel to e^{iϑ_∞} appears after the proof of existence of v_∞ and ϑ_∞; the structure is confusing and should be reorganized. This is a presentation issue, not a mathematical error.","section":"Lemma 7.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-motivated and the overall strategy is plausible, but the current text has several nontrivial gaps in exactly the places where the main theorems depend on compactness and limit-transfer arguments. In particular, Proposition 6.3 is the foundation of the coding theorem, and its proof is not complete. I would not recommend acceptance until these arguments are written out rigorously. The reliance on the authors' own [5] for the approximation lemma is acceptable in principle, but the omitted proof should be included or at least precisely stated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about arXiv:2607.20099. First, the main results are genuinely new: they extend the variational symbolic-dynamics scheme to the critical 1/r^2 two-centre system (without Levi-Civita regularization) and obtain coding, scattering, and trapping for the relativistic problem with perturbations. Second, the paper as written is not ready: two load-bearing steps in the proof are not actually proved, and the curvature remark has algebraic errors. The theorems may well be true, but a referee would need substantial additions.\n\nWhat is new and good: the reduction from the relativistic to the classical strong-force problem was already in [12,10], but the treatment of the critical exponent 2 with this machinery, plus the new scattering and trapping theorems, is a real contribution. The compactness and asymptotic lemmas in Section 7 are carefully done. Borrowing the topological and approximation machinery from [5] is legitimate—the paper is just not self-contained, which is acceptable if [5] is solid.\n\nThe soft spots are in proportion: the biggest is Proposition 6.3. The proof of existence of collision minimisers ends with the claim that minimisers enter a fixed small ball around the centre in uniform time. The justification given—\"there exists at least a point of ∂B_r at infinite distance from q\"—is not an argument. Without a uniform-time bound, the diagonal limit may not converge to the centre, and the curves Γ1, Γ2 that define the itinerary are not established. I believe the claim is true and can be fixed by comparing with a path that goes from q to a boundary point and then radially inward; that gives a length bound logarithmic in 1/|q_n|, which controls the exit time. But the text does not do that.\n\nSecond, Lemma 5.3 (confinement of all other minimisers to D) only treats minimisers that intersect ∂D. A minimiser lying entirely outside D would belong to [(α1α2)^m], and that case is not ruled out. This lemma is used to keep the approximating orbits in a fixed compact region, so it matters.\n\nThird, Remark 6.7: the curvature formula has at least two algebraic slips—the h' term should be 1/|q-c|^3, and the denominator of R(q) has a typo. The sign of the curvature likely survives correction, but the conjugation claim for R=0 is not yet justified.\n\nOverall: the paper is for people working on singular variational methods and celestial mechanics. It deserves peer review, but only with a strong request for revision. I would not cite it in its current form.","headline":"Genuinely new results on critical two-centre coding, but the proof has two load-bearing gaps—especially the uniform-time claim in Proposition 6.3—so the paper needs major revision before it is citable.","tokens_in":30806,"tokens_out":19644,"would_cite":false,"duration_ms":171007,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["70F10","34C28","70G75","37B10","37N05","70H40"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the relativistic two-centre problem contains a compact invariant set whose first-return map is semi-conjugated to a chaotic subshift of finite type, with every admissible crossing sequence realised by a collision-free","keywords":["two-centre problem","relativistic dynamics","Maupertuis functional","strong-force singularity","symbolic dynamics","chaotic subshift","scattering trajectories","variational methods"],"falsifier":"For a high-energy classical reduction satisfying (11), compute the [α1α2] minimiser φ and a minimiser γ in another admissible class such as [α1 α2 α1]; if γ has any point outside the disk D bounded by φ, or if φ and γ intersect exactly once, Lemma 5.3 is false and the coding theorem's disk-confinement collapses.","tokens_in":29862,"feed_emoji":"🌀","tokens_out":6889,"duration_ms":64356,"temperature":0.7,"pith_summary":"The paper shows that the special-relativistic two-centre problem can be reparametrised in time so that it becomes a classical two-centre system with inverse-square strong-force singularities. At energies above the infimum of the effective potential, every admissible symbolic itinerary of crossings between two distinguished curves is realised by a collision-free solution, including bounded, scattering, and one-sided trapped orbits. The bounded trajectories form a compact invariant set on which the first-return map is semi-conjugated—and, for the unperturbed case, conjugated—to a chaotic subshift of finite type. A sympathetic reader should care because this turns a non-integrable relativistic model into a variational problem where complex motions can be prescribed combinatorially.","feed_headline":"Relativistic two-centre problem is provably chaotic","feed_subtitle":"A time-change turns the relativistic system into a classical singular one, where every admissible itinerary is realised.","key_machinery":"Carrying object: the Maupertuis functional M_h(γ)=(1/2∫|γ̇|²)(∫[h+V(γ)]), minimised in homotopy classes of the twice-punctured plane; critical points are exactly energy-fixed solutions. The relativistic surface is rewritten as a classical Hamiltonian with potential (h_rel+V_rel)²/2mc², inverse-square near each centre, so minimisers are collision-free and taut-loop geometry controls them. Two collision minimisers ψ1, ψ2 asymptotic to the centres serve as transversal sections; the itinerary map records signed crossings. Compactness confines all periodic minimisers to one disk per k, and periodic minimisers of growing period converge to the coded bi-infinite orbits.","core_discovery":"The central discovery is that the relativistic two-centre flow, after a time-reparametrisation, is equivalent to a generalised classical two-centre problem with critical inverse-square strong-force singularities. In that system, every admissible sequence over {±1,±2}—no run longer than k, no (±1,∓1) or (±2,∓2) patterns—is realised by a collision-free trajectory whose signed crossings of two curves Γ1, Γ2 reproduce the sequence. Bounded orbits form a compact invariant set on which the first-return map is semi-conjugated to a subshift of finite type, and the semi-conjugation is a conjugation when R=0. The same coding yields scattering solutions with prescribed asymptotic directions and trapped","pith_inferences":["Beyond the paper: the same time-reparametrisation turns any relativistic N-centre superposition into a sum of inverse-square singularities, so the coding approach likely extends to N≥3 with a 2N-symbol alphabet; the authors do not state this extension.","The conjugation for R=0 rests on negative curvature of the Jacobi-Maupertuis metric. The curvature formula in Remark 6.7 offers a testable sign condition for perturbations; checking it numerically would predict when semi-conjugation upgrades to conjugation.","The uniform-distance result suggests the chaotic set persists under slightly weaker singularities (exponent just below 2) but breaks down before the Coulomb exponent 1; a numerical continuation in the singularity exponent would locate that breakdown. This is an extrapolation, not a claim of the paper."],"forward_implications":["If the coding theorem is correct, the relativistic two-centre problem at admissible energies has positive topological entropy and contains a Cantor set of bounded orbits realising every k-admissible itinerary.","For R=0, the conjugation means distinct itineraries correspond to distinct trajectories, so the invariant set is faithfully described by the subshift rather than merely shadowed by it.","The scattering result constructs collision-free trajectories with any finite word and any incoming and outgoing asymptotic directions, so the interaction region behaves as a deterministic scatterer with prescribed symbolic winding.","Trapped orbits interpolate between the two regimes: asymptotically free in the past and exhibiting full one-sided symbolic complexity in the future.","Because the classical reduction is proven equivalent, all these existence statements transfer to the original relativistic equations after the time-change.","The same coding holds for the classical two-centre reduction, giving a unified description of the relativistic and classical models."],"fun_headline_variants":["Chaos proved in relativistic two-centre problem","Time trick reveals chaos in relativistic two-centre system","Relativistic two-centre chaos: every orbit code is realisable","How a time change makes relativistic chaos provable","Symbolic coding tames chaotic relativistic motion"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The coding theorem rests on two partially verified geometric facts: the collision minimisers ψ1 and ψ2 can be chosen as non-intersecting transversal curves, and every periodic minimiser outside the class [α1α2] stays inside the disk D bounded by the [α1α2] minimiser; the paper asserts the latter in Lemma 5.3 without proving that every such minimiser must meet ∂D, and Proposition 6.3 states without proof that minimisers enter every small ball around a centre in uniform time—if","fun_headline_variants_meta":{"raw":{"variants":["Chaos proved in relativistic two-centre problem","Time trick reveals chaos in relativistic two-centre system","Relativistic two-centre chaos: every orbit code is realisable","How a time change makes relativistic chaos provable","Symbolic coding tames chaotic relativistic motion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1104,"prompt_tokens":645,"completion_tokens":459,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":389,"completion_tokens_details":{"reasoning_tokens":383}},"tokens_in":389,"tokens_out":459,"duration_ms":4740,"temperature":1.0,"reasoning_tokens":383,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T10:47:17.683228+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a high-energy classical reduction satisfying (11), compute the [α1α2] minimiser φ and a minimiser γ in another admissible class such as [α1 α2 α1]; if γ has any point outside the disk D bounded by φ, or if φ and γ intersect exactly once, Lemma 5.3 is false and the coding theorem's disk-confinement collapses.","supporting_citations":[],"review_version":1}