{"id":"392beec0-7cdd-44d1-a933-a47d483fd81f","arxiv_id":"2508.15992","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For almost all choices of interaction parameters, the empirical occupation measures of interacting vertex-reinforced random walks converge almost surely to a fixed point of the transition map, whose support structure is characterized by a system of linear equations.","lead":"This paper introduces a general model of several random walkers that reinforce one another on shared vertices and proves that their long-run visit proportions converge to a fixed point for almost all parameter choices. The framework handles cooperative and competitive interactions on arbitrary complete subgraphs, and is illustrated with complete, star, and cycle graphs.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1(iv) quantifies det Rρ(S) over all S∈V, including empty S where D(S)=∅ and det=0, so the hypothesis is never satisfied; Corollary 1's genericity proof needs a restricted S and a non-identical-zero determinant argument.","rationale":"The reader's verdict is CONDITIONAL, and our concern does not move it: the central Lyapunov/stochastic-approximation argument appears sound, and the issue is a fixable quantification over supports. However, the reader's weakest-assumption statement (symmetry of ρ) is not the place where the proof is least secure. The real soft spot is the determinant condition in Theorem 1(iv)/Corollary 1: as written it ranges over all S∈V, including empty supports that are structurally singular, so the hypothesis is never true. This directly affects the paper's headline claim that convergence to a fixed point is generic. The paper's genericity proof cites Lebesgue nullity of singular matrices in the full matrix space, but Rρ(S) is a structured linear image of ρ; nullity of the preimage requires the determinant polynomial to be nonzero, which is not shown. The proposed test settles whether the fix is purely cosmetic or reveals a genuinely nongeneric family of parameters. The example sections (e.g., Theorem 6) also fail to verify the determinant condition before applying Theorem 1(iv), matching the reader's concern, but that is downstream of the same issue. I therefore recommend keeping the CONDITIONAL verdict and asking the authors to clarify the quantification over S and supply the missing genericity argument.","tokens_in":38339,"tokens_out":19139,"duration_ms":212941,"concrete_test":"Take m=1, V^1={1,2}, S=(∅). Equation (10) reads 0=1, so D(S)=∅ and any assigned coefficient matrix has a zero row, giving det=0. This shows the stated hypothesis 'all S∈V' fails at every parameter. Then test the intended fix: for m=2 on K_2 with S_1=S_2={1,2}, ρ12_v=ρ21_v=−1, ρ11=ρ22=−ϵ, compute det Rρ(S); it is nonzero for small ϵ>0. More generally, use symbolic determinant to check that for each nonempty S the determinant polynomial in the free ρ variables is not identically zero; if any nonempty S is identically singular, the 'almost all parameters' claim would fail for that geometry.","verdict_should_be":"UNCHANGED","load_bearing_attack":"V is defined as ℘(V^1)×···×℘(V^m), so it contains S with S_i=∅. For such S, system (10) includes ∑_{ℓ∈∅} x^i_ℓ = 1, i.e., 0=1; hence D(S)=∅ and the coefficient matrix Rρ(S) is either non-square or has a zero row, so det Rρ(S)=0 for every ρ. Consequently the hypothesis of Theorem 1(iv), 'det Rρ(S) ≠ 0 for all S∈V', is never satisfied, making the theorem as stated vacuous. The proof of Corollary 1 relies on this hypothesis, arguing that the set of singular matrices in M_{d,d}(R) is Lebesgue null, and concludes det Rρ(S) ≠ 0 for almost all ρ 'for each S∈V'. For empty S this is false; the generic-convergence claim is therefore not established by the given argument. The intended fix is to quantify over S with D(S)≠∅ (equivalently all S_i nonempty) and to prove for each such S that det Rρ(S), as a polynomial in the symmetric ρ variables, is not identically zero. The paper only cites nullity of singular matrices in the ambient space, which does not imply nullity of the preimage under the structured linear map ρ↦Rρ(S). The same gap propagates to Theorem 6, which invokes Theorem 1(iv) after asserting finiteness of Fix(π) without checking the determinant condition.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a general model of m interacting vertex-reinforced random walks, each confined to a complete subgraph of a common graph, with transition probabilities depending on empirical occupation proportions via parameters η, ρ, and α. The main results describe Fix(π) as a finite union of connected sets, prove that the empirical occupation process converges a.s. to the chain-recurrent set/fixed points under certain regularity conditions, and claim generic a.s. convergence to a single fixed point. The authors also prove non-convergence criteria for boundary and interior points and apply them to complete graphs, stars, and cycles. The proofs are built on stochastic approximation theory, a strict Lyapunov function, and explicit fixed-point classifications.","tokens_in":38717,"tokens_out":9368,"duration_ms":100799,"significance":"If the central claims can be made fully correct, the paper would be a substantial contribution to the interacting reinforced random walk literature, generalizing and unifying prior work on two-walk and complete-graph models. The Lyapunov-function construction (Theorem 10), the stochastic approximation setup, and the explicit classifications for complete graphs, stars, and cycles are valuable and go beyond existing results. The paper is self-contained modulo standard external theorems and does not rely on fitted parameters or simulations. However, several load-bearing points, especially the statement and proof of the generic convergence result, are currently not established, so the significance is conditional on a successful revision.","major_comments":[{"comment":"The quantifier 'for all S∈V' is never satisfiable. V contains tuples S=(S1,…,Sm) with some Si=∅. For such S, system (10) includes ∑_{ℓ∈∅} x^i_ℓ = 1, so D(S)=∅ and the coefficient matrix Rρ(S) has a zero row. Hence det Rρ(S)=0 for every ρ, and Theorem 1(iv) is vacuous. The hypothesis should be restricted to S with all Si nonempty (equivalently D(S)≠∅). This issue propagates to Corollary 1 and Theorem 6.","section":"Section 2, Theorem 1(iv), Eq. (10)"},{"comment":"The proof asserts that f:ρ↦Rρ(S) is a bijection onto M_{d,d}(R) and then uses that the singular matrices form a Lebesgue-null set. This is not valid: Rρ(S) is a linear map into a proper subspace of M_{d,d}(R), and when some Si=∅ it is even a constant map with a zero row. Nullity of the singular set in the ambient matrix space does not imply nullity of its preimage under a non-surjective linear map. One must prove directly that det Rρ(S) is not identically zero as a polynomial in ρ for each relevant S, e.g. by exhibiting one parameter choice for which system (10) has a unique solution. As written, the generic convergence claim is unproved.","section":"Section 4.2, proof of Corollary 1"},{"comment":"The proof invokes 'items (iii) and (iv) of Theorem 1' to pass from P(L(X)⊂Fix_U)=0 for U≠∅ to P(L(X)⊂Fix_∅)=1. But Theorem 1(iv) is not applicable in this ϵ=0 complete-graph regime: for supports Si with |Si|≥2 and disjoint supports, the coefficient matrix Rρ(S) has zero rows (ρ_ii=0 and the other walk is absent on those vertices), so det Rρ(S)=0. The conclusion may still follow from Corollary 3/Theorem 2, which the authors prove and which applies here, but it is not obtained by the argument given in the proof of Theorem 5.","section":"Section 6, proof of Theorem 5"},{"comment":"Theorem 6 applies Theorem 1(iv) after Lemma 9, but no verification of det Rρ(S)≠0 is supplied for the supports contributing to Fix(π)=˜K∪˜Kc1∪˜Kc2. Since Theorem 1(iv) as stated is vacuous, and even the corrected version requires the determinant condition, the conclusion 'X converges almost surely to a single point' is not established for the complete-graph example. The same unverified invocation appears in Theorem 7 for ϵ>0. The authors need either a direct determinant computation for each support in these examples or an alternative argument (e.g., using Theorem 2/Corollary 3 when applicable) to justify L(X)⊂Fix(π).","section":"Section 6, proof of Theorem 6; also Theorem 7"}],"minor_comments":[{"comment":"The displayed formula π_i^v(x)= x_i^v(∂L(x)/∂x_i^v)^α / N^i(x) has the wrong sign: since ∂L/∂x_i^v <0, the right-hand side is negative for α=1 and non-real for non-integer α. It should be x_i^v(-∂L(x)/∂x_i^v)^α / N^i(x). The subsequent algebra appears consistent with the sign-corrected version, so this is likely a typo.","section":"Proof of Theorem 10, Step 1"},{"comment":"The statement is about 'almost all (α,η,ρ)', but the proof only varies ρ and treats η and α as fixed. This can be repaired by a Fubini/coarea argument, but as written the parameter space dimension is not handled.","section":"Proof of Corollary 1"},{"comment":"There are several typos and encoding artifacts: 'vetor field' instead of 'vector field', 'T¨oeplitz' instead of 'Toeplitz', 'Departameto' instead of 'Departamento', and occasional missing spaces around citations. These should be cleaned up.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The vacuity of Theorem 1(iv) appears to be an oversight, but it is load-bearing: Corollary 1 and the complete-graph applications rest on it. I would insist that the authors (1) restate the determinant condition only for supports with nonempty coordinates, (2) prove or disprove that det Rρ(S) is not identically zero for each such S, and (3) rewrite the proofs of Theorems 5–7 so that the convergence-to-Fix step is justified either by the corrected Theorem 1(iv) or by Theorem 2/Corollary 3. The paper has merit and the main architecture seems sound, but these gaps currently block the advertised generic-convergence result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read of Prado–Rosales. The model is new and worth engaging with: m vertex-reinforced walks on complete subgraphs interacting through shared vertices, with real-valued interaction parameters and intrinsic weights. The Lyapunov function in Theorem 10 is a real extension of Benaïm's single-walk construction, the fixed-point characterization (Theorem 1(i)-(ii), Corollary 2) is clean, and the non-convergence criteria (Theorems 3-4) plus the star/cycle analyses look sensible. The stochastic approximation setup (Lemmas 2-3) is standard and appears correctly applied.\n\nThe problem is that the headline result, as stated, is vacuous. Theorem 1(iv) assumes det R_rho(S) ≠ 0 for all S in V. But V is the full product of power sets, so it contains S with S_i = ∅. For such S, system (10) contains sum over empty set of x^i_l = 1, i.e., 0=1; D(S) = ∅ and the coefficient matrix has a zero row, so det R_rho(S) = 0 identically. The hypothesis never holds, so Theorem 1(iv) proves nothing. Corollary 1 inherits this: the proof claims the singular matrices form a null set in the ambient space and pulls that back to parameter space, but for empty S the bad set is all of Theta, and for nonempty S the determinant is a polynomial whose zero set is only null if the polynomial is not identically zero—which the paper never shows. The pullback of a null set under the linear map rho -> R_rho(S) can be everything. The fix is probably local: restrict to S with all S_i nonempty and check det R_rho(S) not-identically-zero case by case, but that work is not in the paper. Theorem 6 and Remark 1 also cite Theorem 1(iv) after proving only that Fix(pi) is finite, which is not the same as checking the determinant condition.\n\nOne more small thing: for |S_i| ≥ 3, the first line of (10) contains redundant equations, so \"the matrix of coefficients\" and its determinant need a clean definition; currently it is ambiguous which submatrix R_rho(S) is.\n\nNet: the model and much of the machinery are solid, and the flaw looks repairable. The paper deserves a serious referee—I would send it out with an explicit request to fix the quantification over S and prove the non-vanishing determinant statements. Not a reject, but the advertised generic-convergence result is currently unproven.","headline":"New interacting reinforcement model with a genuinely useful Lyapunov construction, but Theorem 1(iv) is vacuous as stated and the genericity claim needs repair.","tokens_in":39188,"tokens_out":9185,"would_cite":false,"duration_ms":96217,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60K35","37C10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the joint empirical occupation measure of several interacting vertex-reinforced random walks converges almost surely to the set of fixed points of their transition probabilities, and to a single fixed point for almost","keywords":["reinforced random walk","interacting random walks","stochastic approximation","Lyapunov function","fixed point convergence","occupation measure","competitive dynamics","complete graphs"],"falsifier":"Run the paper's ε = 0 competitive two-walk model on K5 for sufficiently many steps and record the final supports and overlap H(X(n)). Theorem 5 predicts that H(X(n)) → 0 almost surely and that the limit set is contained in the segregated set K of disjoint supports; a single trajectory in which both walks keep positive occupation on a shared vertex with positive probability, or in which the overlap fails to decay to 0, would settle the claim by counterexample.","tokens_in":38240,"feed_emoji":"🎲","tokens_out":8782,"duration_ms":100846,"temperature":0.7,"pith_summary":"This paper studies several random walks that reinforce the vertices they visit, with the twist that each walk's next-step probabilities also depend on how often the other walks have visited shared vertices. The central result is a strong law: the vector of cumulative visit proportions converges almost surely to a point where the transition probabilities reproduce themselves, i.e., to a fixed point of the system. For almost every choice of the interaction strengths and intrinsic vertex preferences, convergence is to a single deterministic fixed point rather than merely to a set; the exceptional parameters form a measure-zero set. The proof uses a strict Lyapunov function that decreases along the limiting dynamics unless the process is already at a fixed point, plus non-convergence tests that rule out boundary and linearly unstable limits. Concrete analyses of competing walks on complete graphs, stars, and cycles show which segregated or shared configurations can persist.","feed_headline":"Almost all interacting reinforced walks converge to fixed points","feed_subtitle":"Joint occupation proportions settle on deterministic limits for almost every choice of attractions and repulsions.","key_machinery":"The machine is the strict Lyapunov function L(x) = -Σ η^i_v x^i_v - ½ Σ_v Σ_{i,j∈I_v} ρ^ij_v x^i_v x^j_v, paired with the stochastic approximation representation X(n+1)-X(n) = (F(X(n)) + U(n)) Ξ_n, F = -x + π(x). The gradient inner product factorizes as a sum over walks of g_i(x)⟨φ(x̃^i/π̃^i(x)), x̃^i Γ̃^i(x)⟩, where φ(z) = -z^{-1/α} and Γ̃^i is the generator matrix of a finite Markov chain with invariant measure π̃^i. A spectral-gap inequality bounds this term by -λ Σ_v (x^i_v - π^i_v(x))²/π^i_v(x), strictly negative off Fix(π), turning chain-transitivity of the stochastic approximation limit set into almost-sure convergence to fixed points. The fixed-point linear systems (10), with coeffic","core_discovery":"The joint occupation process X(n) satisfies a stochastic approximation recursion whose mean vector field is F(x) = -x + π(x). The fixed-point set Fix(π) is exactly the finite union of the convex solution sets D(S) of the linear system (10) over support patterns S, and Fix(π) is always nonempty. If the coefficient matrix of every such linear system is nonsingular, Fix(π) is finite and X(n) converges almost surely to one of its points; since singular matrices form a null set, this single-point convergence is generic. The argument constructs the strict Lyapunov function L(x) = -Σ η^i_v x^i_v - ½ Σ ρ^ij_v x^i_v x^j_v, whose derivative along the flow is non-positive and vanishes only on Fix(π), v","pith_inferences":["The paper notes, but does not pursue, the resemblance of the α = 1 transition probabilities to Lotka-Volterra maps; that connection suggests replicator-equation stability tools could be used to predict which of several stable fixed points is selected from given initial data.","The boundary non-convergence criterion offers a practical screening rule: before solving the linear systems for every support pattern S, one can discard any support whose relative transition odds at a candidate boundary exceed one, sharply pruning the exponential list of patterns.","When det R_ρ(S) = 0, the fixed-point components are continua and the paper leaves open whether the empirical process locks onto a random point of the continuum or wanders along it; a focused simulation study on the K3 example with η = ρ^ii = 0 could distinguish these alternatives.","The generic single-point convergence result is measured in Lebesgue measure on parameter space, so exactly symmetric regimes such as ε = 0 are exceptional null sets; this suggests that symmetry tuning is precisely where qualitatively different, non-point-convergent behavior can be observed experimentally."],"forward_implications":["On complete graphs with two competitive walks, ε = 0 yields almost-sure eventual segregation onto disjoint vertex sets, so the overlap H(X(n)) tends to 0 almost surely; for small ε > 0 the walks converge to explicit fixed points with overlap below κ³ε².","On star graphs, the competitive dynamics has a phase transition at ε = 1/2: for ε = 0 one walk dominates the center; for 0 < ε < 1/2 a subset of walks shares the center at a frequency ε/(|K|+2ε-1); for 1/2 < ε < 1 all walks share the center symmetrically.","On cycles with ε = 0, the limit set is confined to four explicitly described families of fixed points, including a periodic alternating-edge family when m is a multiple of four; with small ε > 0 the process converges to a single fixed point.","Any boundary fixed point whose limiting relative transition probability to a zero-occupation vertex exceeds one is almost surely unattainable, and any interior fixed point that is linearly unstable for the vector field is likewise almost surely avoided.","For almost all parameter values satisfying the symmetry and positivity assumptions, the joint occupation vector converges almost surely to a single point of Fix(π), not merely to the fixed-point set."],"supporting_citations":[{"why":"Supplies the stochastic approximation limit-set lemma and the chain-recurrence plus strict Lyapunov function lemma that turn descent into almost-sure convergence to Fix(π).","marker":"[2]"},{"why":"Supplies the spectral-gap inequality used to prove the gradient inner product is strictly negative away from fixed points.","marker":"[5]"},{"why":"Supplies the linearly unstable equilibrium non-convergence criterion adapted as Theorem 11, used to rule out interior fixed points.","marker":"[18]"},{"why":"Provides the original single-walk vertex-reinforced model being generalized and the boundary non-convergence argument adapted for Theorem 3.","marker":"[19]"},{"why":"The closest prior model of two repelling walks on complete graphs, whose overlap results are sharpened by the new convergence and limit-point characterization.","marker":"[9]"},{"why":"Prior interacting-walk convergence under isolated limit sets, and its Lemma 9 helps verify the noise condition used in the linearly unstable non-convergence argument.","marker":"[22]"},{"why":"Origin of the relative-entropy Lyapunov construction that the paper adapts to the interacting multi-walk setting.","marker":"[8]"},{"why":"Provides the characteristic polynomials of tridiagonal Toeplitz matrices used to show cycle fixed-point sets are finite for small ε in Theorem 9.","marker":"[7]"}],"fun_headline_variants":["Reinforced walks on subgraphs converge almost surely","Almost sure limits for interacting reinforced walks","Generic convergence in vertex-reinforced random walks","Interacting walks settle on fixed points generically","Attraction and repulsion still lead to convergence"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The symmetry condition ρ^ij_v = ρ^ji_v is load-bearing: the strict Lyapunov descent and gradient inequality use the symmetric quadratic overlap term, so without this mirror symmetry the almost-sure convergence to Fix(π) is not established.","fun_headline_variants_meta":{"raw":{"variants":["Reinforced walks on subgraphs converge almost surely","Almost sure limits for interacting reinforced walks","Generic convergence in vertex-reinforced random walks","Interacting walks settle on fixed points generically","Attraction and repulsion still lead to convergence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000209,"raw_usage":{"total_tokens":1283,"prompt_tokens":822,"completion_tokens":461,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":566,"completion_tokens_details":{"reasoning_tokens":391}},"tokens_in":566,"tokens_out":461,"duration_ms":5679,"temperature":1.0,"reasoning_tokens":391,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T17:36:56.448396+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the paper's ε = 0 competitive two-walk model on K5 for sufficiently many steps and record the final supports and overlap H(X(n)). Theorem 5 predicts that H(X(n)) → 0 almost surely and that the limit set is contained in the segregated set K of disjoint supports; a single trajectory in which both walks keep positive occupation on a shared vertex with positive probability, or in which the overlap fails to decay to 0, would settle the claim by counterexample.","supporting_citations":[{"cited_title":"Control Optim","cited_arxiv_id":null,"evidence_quote":"Supplies the stochastic approximation limit-set lemma and the chain-recurrence plus strict Lyapunov function lemma that turn descent into almost-sure convergence to Fix(π)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the spectral-gap inequality used to prove the gradient inner product is strictly negative away from fixed points."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the linearly unstable equilibrium non-convergence criterion adapted as Theorem 11, used to rule out interior fixed points."},{"cited_title":"Theory Related Fields 92 (1992), no","cited_arxiv_id":null,"evidence_quote":"Provides the original single-walk vertex-reinforced model being generalized and the boundary non-convergence argument adapted for Theorem 3."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The closest prior model of two repelling walks on complete graphs, whose overlap results are sharpened by the new convergence and limit-point characterization."},{"cited_title":"Rosales, Fernando P","cited_arxiv_id":null,"evidence_quote":"Prior interacting-walk convergence under isolated limit sets, and its Lemma 9 helps verify the noise condition used in the linearly unstable non-convergence argument."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Origin of the relative-entropy Lyapunov construction that the paper adapts to the interacting multi-walk setting."},{"cited_title":"Grudsky,Spectral properties of banded Toeplitz matrices, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2005","cited_arxiv_id":null,"evidence_quote":"Provides the characteristic polynomials of tridiagonal Toeplitz matrices used to show cycle fixed-point sets are finite for small ε in Theorem 9."}],"review_version":1}