{"id":"107c57da-01c0-430d-a78c-570296fc2ed2","arxiv_id":"2607.23640","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Coupled density fluctuations of a two-species SSEP with critical slow reservoirs converge to a generalized Ornstein–Uhlenbeck process with bulk, conversion, and boundary noise.","lead":"The paper proves that density fluctuations in a two-species exclusion process with slow open boundaries converge to a generalized Ornstein–Uhlenbeck process. It extends one-species slow-boundary fluctuation theory to coupled multi-component systems with bulk conversion and Robin boundary noise.","discovery_kind":"extension","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The reader's soft spot is the right one: Theorem 2.1 (hydrodynamic limit) is asserted, not proved, and it bites at exactly one place — Proposition 5.1's identification of Q_t's coefficients. The decisive question is whether the species-resolved limit closes triangularly from the known one-species慢…","rationale":"I independently re-checked the places where the conclusion could break and found the reader's weakest assumption to be the correct load-bearing one. What survives scrutiny: (1) the boundary matrices K_L, K_R and conversion matrix M in (2.1) match a direct computation from the microscopic generator, so the asserted PDE is the right one; (2) the carré du champ computations (5.2)–(5.3) are consistent with the claimed limiting form Q = Q^s + Q^c + Q^{bd}, including signs and the off-diagonal −2ρ_1ρ_2∂f_1∂f_2 cross term; (3) the boundary quadratic variation identification needs only time-averaged boundary occupations, which concentrate since the boundary site mixes at rate O(n) — standard and fine; (4) the genuinely novel technical piece, the two-point correlation bound of Proposition 8.3 via the nine-component Markovian extension (needed because the four-component boundary matrices have negative off-diagonal entries), is proved in-house with a plausible domination by an absorbed scalar walk, using only a cited occupation-time estimate from [3, Lemma 3.2]; (5) tightness and OU uniqueness follow standard, correctly-applied machinery (Mitoma, Aldous, analytic semigroups). What does not survive as \"proved\" is Theorem 2.1 itself, and I sharpened the reader's concern to its exact point of application (Proposition 5.1 only) and to a checkable algebraic question — triangular closure of the color-difference field given the autonomous total density, the latter being exactly the known one-species result of [3]. This makes the missing proof look likely to succeed, but \"likely\" is why the verdict is CONDITIONAL rather than ACCEPT, and nothing I found pushes toward REJECT. Hence UNCHANGED, agreeing with the reader.","tokens_in":38595,"tokens_out":8817,"duration_ms":326294,"concrete_test":"Check triangular closure of the hydrodynamic system. From (2.4), derive the discrete evolution of ξ^n(x,t) = γ_1ρ^{n,t}_1(x) − γ_2ρ^{n,t}_2(x): in the bulk it is autonomous (n²Δξ − (γ_1+γ_2)ξ); at x=1 confirm the boundary equation is ∂_tξ(1) = n²(ξ(2)−ξ(1)) + n[q_0ρ^{n,t}_0(1) − α_0ξ(1)] with q_0 = γ_1α_1 − γ_2α_2, which involves only the vacancy ρ_0 = 1 − (ρ_1+ρ_2). If this closure holds exactly, Theorem 2.1 follows from [3] (for U) plus a standard one-block argument for ξ, and the verdict can move toward ACCEPT. If cross-species two-point correlations enter the boundary or bulk equations for ξ, the entropy-method adaptation requires genuinely new estimates and the CONDITIONAL verdict must stand or weaken.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The centering in (2.5) uses the exact microscopic expectation ρ^{n,t}_k(x), so the drift cancellation in §5 — including the boundary cancellation (5.6), which rests only on the adjoint Robin condition on g ∈ S† plus Taylor expansion — is exact and does not use Theorem 2.1. The unproved hydrodynamic limit is load-bearing at exactly one point: Proposition 5.1, where local block averages (via Lemma 4.2) and boundary-site time averages must converge to ρ(s,u) and its traces ρ(s,0), ρ(s,1) solving (2.1) with the specific matrices K_L, K_R. All three noise components Q^s, Q^c, Q^{bd} inherit their coefficients from this. This is also where an \"adaptation of [3, 21]\" is most likely to conceal real work: with unequal reservoirs there is no product invariant measure, and the slow-boundary two-blocks estimates of [3] must be redone for the coupled boundary generator L^±_n. Two mitigating facts lower the risk that the stated limit is wrong (as opposed to unproved): (i) I verified K_L, K_R against the microscopic rates — e.g. injection at α_k/n, removal at α_0/n gives ∂_uρ_1(0) = α_1 − (1−α_2)ρ_1 − α_1ρ_2, matching K_L exactly, and summing components yields the scalar Robin condition ∂_uU = U − (α_1+α_2) for U = ρ_1+ρ_2; (ii) the total occupation η_1+η_2 evolves autonomously as the one-species slow-boundary SSEP of [3] with parameters α_1+α_2, α_0, so half of Theorem 2.1 is already a known theorem, not an adaptation. The genuinely new content is the species-resolved (color-difference) field, whose closure needs checking.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The paper studies a two-species symmetric simple exclusion process on {1,…,n−1} with bulk species conversion (rates γ1, γ2) and slow boundary reservoirs (rates of order 1/n), in the critical regime producing Robin boundary conditions macroscopically. The main results (Theorems 2.2–2.3) state that the coupled density fluctuation field, centered at the exact microscopic expectations ρ^{n,t}_k(x), converges in D([0,T],(S†)') to a generalized Ornstein–Uhlenbeck process with drift A† = ΔI₂ + Mᵀ and quadratic covariation ∫Q_s ds, where Q = Q^s + Q^c + Q^{bd} decomposes into bulk-exchange, conversion, and boundary-reservoir contributions. The proof develops: analytic semigroups for A and A† on Robin domains with exponential decay (Prop. 3.1, Lemma 3.1); Dirichlet-form estimates and a replacement lemma (Lemmas 4.1–4.2); quadratic-variation identification (Prop. 5.1); uniqueness of the martingale problem (Prop. 6.1); tightness via Mitoma/Aldous (§7); and two-point correlation estimates via a nine-component Markovian enlargement including vacancies (§8, Props. 8.1–8.4).","tokens_in":39049,"tokens_out":3650,"duration_ms":409953,"significance":"If correct, this is a solid extension of the one-species slow-boundary fluctuation theory of Franco–Gonçalves–Neumann and Gonçalves–Jara–Menezes–Neumann to a genuinely coupled two-species setting, and it is the first fluctuation result combining slow (Robin) boundaries with bulk species conversion under a shared exclusion constraint. Two technical contributions deserve explicit credit: (i) the nine-component correlation process on V_n × {0,1,2}² (§8.2), which restores a Markov-generator structure (nonnegative off-diagonal rates, zero column sums) that the four-component system lacks, allowing the diagonal occupation-time estimate of Baldasso–Menezes–Neumann–Souza to be imported — this is a clean answer to a real obstruction the authors correctly identify; (ii) the explicit verification that the boundary matrices K_L, K_R in (2.1) match the microscopic reservoir rates (I checked: injection at α_k/n, removal at α_0/n gives ∂_uρ_1(0) = α_1 − (1−α_2)ρ_1 − α_1ρ_2, and summing components recovers the scalar Robin condition ∂_uU = U − (α_1+α_2) for U = ρ_1+ρ_2). The result is a pure scaling-limit theorem with no fitted parameters, and the limiting covariance σ(S_t f, S_s g) + ∫Q_r dr is a","major_comments":[{"comment":"The hydrodynamic limit is stated but not proved ('It can be obtained by adapting the standard entropy method… as in [3, 21]'). This is load-bearing at exactly one point: Proposition 5.1, where block averages (via Lemma 4.2) and boundary-site averages must converge to ρ(s,u) and its traces solving (2.1) with the specific matrices K_L, K_R; all three components of Q inherit their coefficients from this. The drift cancellation in §5 (including the boundary cancellation (5.6)) uses only the exact microscopic centering and the adjoint Robin condition, so it does not depend on Theorem 2.1 — but the noise identification does. Two facts lower the risk that the stated limit is wrong (as opposed to unproved): the total occupation η_1+η_2 evolves autonomously as the one-species slow-boundary SSEP of [3] with parameters α_1+α_2, α_0, so half of Theorem 2.1 is already a theorem; and the K_L, K_R matr","section":"§2.2, Theorem 2.1; used in Prop. 5.1"},{"comment":"The identification of the boundary terms in the quadratic variation — the [α_k η^s_0(1) + α_0 η^s_k(1)]f_k(1/n)² terms in (5.2) converging to (α_kρ_0(s,0)+α_0ρ_k(s,0))f_k(0)² — is dismissed with 'The remaining items can also be obtained using a similar method.' This is not a similar method: Lemma 4.2 replaces occupation variables by block averages over boxes contained in Σ_n, and identifying the limit of a single boundary site's time average with the trace ρ(s,0) requires a boundary replacement (two-blocks) estimate pinned to the reservoir densities, which in the slow-boundary literature is the technically delicate step (cf. [3, §5]). As written, the boundary parts of Φ_kk, and hence Q^{bd}, are not derived. A proof, or a precise citation of a lemma that covers time-averaged boundary-site occupation for this coupled boundary generator, is needed.","section":"§5, proof of Proposition 5.1 (final paragraph)"},{"comment":"The uniqueness argument uses 'the limiting field satisfies Y ∈ C([0,T],(S†)')' to conclude uniform continuity of (s,r) ↦ Y_s(A†S_r f). But tightness in §7 is proved in D([0,T],(S†)'), and no argument is given that limit points are supported on continuous paths. The standard route (vanishing jumps of Y^n_t(f), uniformly O(n^{-1/2}), plus continuity of the limiting quadratic variation) is available and the jumps are indeed bounded by C‖f‖_∞/√n as noted in Prop. 5.1, but the step must be made explicitly, since the uniqueness proof consumes it.","section":"§6.1, proof of Proposition 6.1"}],"minor_comments":[{"comment":"The symbol P is used for two different matrices (the 4×4 diagonal exchange matrix in §8.1 and the 9×9 one in §8.2), and φ^{n,t}(x,y) denotes both the four-component and nine-component vectors. Distinct notation would prevent confusion, especially since both appear in Propositions 8.1 and 8.2.","section":"§8.1–8.2"},{"comment":"In the definition of N^{n,t}_{k,l}, Γ_n(Y^{n,s}_k(f_k), Y^{n,s}_l(f_l)) is later written as Γ_n(f_k, f_l) in the bracket ⟨M^n_k(f_k), M^n_l(f_l)⟩_t; the shorthand should be introduced or harmonized.","section":"§5, below (5.1)"},{"comment":"The term R^n_0(s) arises because the bulk conversion sum runs over x ∈ {2,…,n−2} while the fluctuation field includes x=1 and x=n−1; a one-line explanation of this bookkeeping would help the reader, since the O(n^{-1/2}) bound relies on it.","section":"§5, (5.5)"},{"comment":"The phrase 'Doob martingale' (§5, first page) presumably means the Doob–Meyer decomposition / Dynkin martingale; standard terminology would be clearer.","section":"§2.1"},{"comment":"Reference [1] (Aldous) is cited for the Fréchet-space property of C^∞([0,1]) in Prop. 7.1, which is not the content of that paper; a standard functional-analysis reference (e.g., Treves or Rudin) is appropriate. Similarly, in §6.1 'Ito's Formula (see [[22], Theorem 3.3 and Corollary 3.3])' has a stray double bracket and the specific result in Revuz–Yor should be pinpointed.","section":"References"},{"comment":"In the second-moment computation for Y^n_0(f), the expression mixes χ(ρ^{n,0}_k(x)) with the cross term −2ρ_1ρ_2; defining χ as the full 2×2 susceptibility matrix (as in §2.3.2) rather than the scalar χ(r)=r(1−r) would make the two displays consistent.","section":"§7"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a competent, conventionally structured contribution in the Franceschini–Gonçalves–Jara / Franco–Gonçalves–Neumann line, and the nine-component correlation device in §8 is a genuine technical addition. The two load-bearing gaps (unproved Theorem 2.1; boundary terms in Prop. 5.1) are, in my assessment, closable within the paper's scope — the centering trick makes the drift exact, and the autonomous total-occupation dynamics anchors half of the hydrodynamics — but they are proof gaps, not presentation issues, hence major rather than minor revision. Fit with the journal's scope is good assuming it publishes scaling-limit papers in interacting particle systems."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The new piece is clear: non-equilibrium fluctuations for a two-species SSEP with critical slow boundaries (Robin) and bulk conversion, with the coupled field going to a generalized OU whose quadratic form splits into bulk exchange, conversion, and boundary reservoir noise. One-species slow-boundary fluctuations and multi-species open hydro already exist; this combination does not, and the authors set it up cleanly.\n\nWhat they do well is the full fluctuation architecture. Analytic semigroups for the Robin adjoint, Dirichlet-form estimates, a usable replacement lemma, Dynkin martingales with explicit carré-du-champ, OU uniqueness via the usual complex exponential martingale, Mitoma/Aldous tightness, and especially the nine-component correlation system that restores a Markov generator when the four-component boundary matrices lose non-negative off-diagonals. The adjoint Robin cancellation at the boundary (their (5.6)) is exact and only needs Taylor plus the test-function space. Centering is on the microscopic means, so the drift identity does not secretly rely on the continuum profile.\n\nThe soft spot the reader and stress-test flag is real and correctly localized. Theorem 2.1 is asserted by “adaptation” of Baldasso et al. and Mourragui–Saada–Velasco; it is not proved. That bites almost only at Proposition 5.1, where block averages and boundary occupation must converge to the continuum ρ that supplies the coefficients of Q^s, Q^c, and Q^bd. Mitigating facts: total occupation η1+η2 is exactly the one-species slow-boundary SSEP of [3], so half the hydro is already a theorem, and the matrices KL, KR match the microscopic rates by direct check. Still, for a complete paper the species-resolved hydro (and the two-block estimates under the coupled L±n) should be written or cited as a finished lemma, not waved through.\n\nNo circularity, no free parameters, citation pattern is appropriate. This is for people who already work on open IPS fluctuations; they will get a usable martingale-problem template and a correlation trick worth stealing. It deserves a serious referee, not a desk reject. I would engage: read the hydro gap carefully, then use the OU characterization if I need multi-species Robin noise.","headline":"Solid multi-species slow-boundary fluctuation theorem; the only real soft spot is an asserted (not proved) hydrodynamic limit that mainly loads the noise coefficients.","tokens_in":40111,"tokens_out":569,"would_cite":true,"duration_ms":23015,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60K35","82C22"],"pacs":[],"model":"grok-4.5","headline":"Two-species exclusion with critical slow boundaries has density fluctuations that converge to a generalized Ornstein–Uhlenbeck process with Robin noise.","keywords":["multi-species exclusion","slow boundary","Robin boundary conditions","non-equilibrium fluctuations","Ornstein-Uhlenbeck process","reaction-diffusion","two-point correlations"],"falsifier":"Compute or simulate the two-point correlation functions of the microscopic process and check whether their scaled version remains O(1/n) uniformly in time; if the bound fails, the second-moment control used for tightness and the identification of Q collapse.","tokens_in":39681,"feed_emoji":"📉","tokens_out":797,"duration_ms":17879,"temperature":0.7,"pith_summary":"This paper studies a one-dimensional lattice where two kinds of particles hop, convert into each other, and cannot occupy the same site, while slow reservoirs at the ends inject and remove particles at rates of order 1/n. At large scale the mean densities obey a reaction–diffusion equation with Robin boundary conditions. The authors prove that the joint microscopic fluctuations around those means, properly scaled, converge to a Gaussian process of Ornstein–Uhlenbeck type. The limiting noise has three pieces: conservative bulk exchange, bulk species conversion, and independent boundary injection/removal. The result supplies a precise fluctuation theory for multi-component open systems driven far from equilibrium by slow reservoirs.","feed_headline":"Two-species slow-boundary exclusion fluctuates as OU noise","feed_subtitle":"Coupled densities converge to a Gaussian process whose noise mixes bulk hops, conversion, and Robin reservoirs","key_machinery":"The linear martingale problem for the adjoint operator A† = ΔI₂ + Mᵀ on the test-function space S† of smooth vector fields satisfying all adjoint Robin compatibility conditions, with quadratic covariation given by the explicit form Q = Qˢ + Qᶜ + Qᵇᵈ.","core_discovery":"Under natural initial assumptions, the coupled fluctuation field of the two empirical densities converges in the Skorokhod space of distribution-valued paths to the unique generalized Ornstein–Uhlenbeck process whose drift is the adjoint linearized Robin reaction–diffusion operator and whose quadratic variation is the sum of bulk exchange, conversion, and boundary reservoir bilinear forms.","pith_inferences":["The nine-component Markovian embedding used for correlations suggests a systematic route to fluctuation proofs for any finite number of species under exclusion and slow boundaries.","Once the missing hydrodynamic limit is supplied, the same martingale problem should yield the stationary fluctuation covariance by sending t → ∞ via the exponential decay of the semigroup.","The composite noise structure (exchange + conversion + boundary) is a concrete prediction that could be checked numerically against particle simulations at moderate n."],"forward_implications":["Non-equilibrium fluctuations for two-species exclusion with critical slow boundaries are fully characterized by a closed Gaussian martingale problem.","Boundary reservoirs contribute additive white-noise terms at the endpoints whose intensities are fixed by the local Robin densities.","Bulk conversion appears as a non-conservative white-noise source proportional to the local conversion rate.","The same framework extends, in principle, to other multi-component open systems whose hydrodynamics yield Robin reaction–diffusion equations."],"fun_headline_variants":["Two-species slow-boundary exclusion converges to generalized OU","Coupled densities yield OU noise from bulk, conversion, Robin edges","Slow reservoirs push two-species exclusion fluctuations to OU","Two-species exclusion with 1/n boundaries fluctuates as OU process","Robin-boundary two-species fluctuations limit to generalized OU"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"The macroscopic density profile that centers the fluctuations is taken as given by a hydrodynamic limit that the paper does not prove, only asserts can be adapted from earlier one-species and multi-species arguments.","fun_headline_variants_meta":{"raw":{"variants":["Two-species slow-boundary exclusion converges to generalized OU","Coupled densities yield OU noise from bulk, conversion, Robin edges","Slow reservoirs push two-species exclusion fluctuations to OU","Two-species exclusion with 1/n boundaries fluctuates as OU process","Robin-boundary two-species fluctuations limit to generalized OU"]},"model":"grok-4.5","effort":"low","cost_usd":0.00381,"raw_usage":{"total_tokens":1086,"prompt_tokens":622,"num_sources_used":0,"completion_tokens":69,"cost_in_usd_ticks":38104000,"prompt_tokens_details":{"text_tokens":622,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":395,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":622,"tokens_out":69,"duration_ms":8369,"temperature":1.0,"reasoning_tokens":395,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T16:50:51.101705+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Compute or simulate the two-point correlation functions of the microscopic process and check whether their scaled version remains O(1/n) uniformly in time; if the bound fails, the second-moment control used for tightness and the identification of Q collapse.","supporting_citations":[],"review_version":1}