{"id":"ae043bb6-8458-4525-b073-9f3246e9705f","arxiv_id":"2607.16411","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A multibath semigroup influence-matrix framework reproduces the two-channel-Kondo critical exponent and Kibble-Zurek scaling in the two-impurity Anderson model.","lead":"This paper extends a tensor-network tool for simulating small systems coupled to many quantum baths, and tests it on two magnetic dots in four baths. It reports that the tool reproduces known critical exponents and captures slow and sudden quench dynamics near a quantum phase transition.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (6) additive-cutoff subtraction is the linchpin of the ν=2 claim; its validity is only supported by an internal collapse and could mask a K-dependent cutoff artifact.","rationale":"The reader identified the additive-cutoff correction in Eq. (6) as the weakest assumption, and I agree. This is the single most load-bearing point because the paper's strongest claim—universal 2CK scaling from real-time dynamics—is directly extracted from the subtraction procedure. The collapse shown in Fig. 2(d) provides some evidence for the additive form, but it cannot rule out a K-dependent cutoff contribution that might distort the exponent precisely in the critical region. The KZ scaling argument in Appendix B rests on the same τ(K), so the v^{2/3} claim also depends on Eq. (6). The method's broader utility as a numerical tool is less threatened: even if the extrapolation is imperfect, the framework could still be practical, but the headline scientific results would be compromised. I see no internal inconsistency or reason to believe the method is fundamentally flawed; rather, the evidence is not airtight enough to promote beyond CONDITIONAL. A clean test—either a sensitivity analysis of Kc or a direct low-T/large-χ run without subtraction—would substantially increase confidence. Since the reader already set CONDITIONAL and my analysis does not move the verdict, I recommend UNCHANGED.","tokens_in":15829,"tokens_out":10045,"duration_ms":107681,"concrete_test":"Test the robustness of the subtracted exponent to the assumed Kc: obtain Kc independently from the static spin-spin correlation ⟨S1·S2⟩_eq(K) crossing (or from NRG), then re-apply Eq. (6) with this Kc and refit the collapsed τ(K) near criticality. If shifting Kc by ±1% (≈±0.009Γ) changes the extracted ν by more than the reported error bars, the ν=2 result is an artifact of the Kc choice rather than robust scaling. Additionally, run one quench set at T=0.001Γ and χ=2000 and check whether the raw τ^{-1}(K) (without subtraction) already vanishes at Kc and follows |K−Kc|^2 over at least a decade; if it does, Eq. (6) is not load-bearing, but if a finite floor persists and only subtraction yields the divergence, the additive form must be independently justified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central result—the universal 2CK exponents τ∼|K−Kc|^−2 and W_d∼v^{2/3}—rests on Eq. (6), which assumes that finite temperature T and SGIM bond dimension χ each contribute a K-independent additive correction to the inverse relaxation rate. This assumption is never derived; it is only 'validated' by the collapse of curves after subtracting τ^{-1}(Kc;T,χ). Collapse is a necessary but not sufficient test: there exist plausible alternative cutoff dependencies (e.g., multiplicative or with K-dependent coefficients) that would also produce an approximate collapse over a finite K-window yet yield a different apparent exponent near Kc. Furthermore, the procedure requires an independent Kc; the paper states Kc≈0.91Γ without a systematic determination or error analysis. If the true cutoff effect is not a K-independent additive constant, subtracting τ^{-1}(Kc;T,χ) would leave a spurious divergence or round it off, biasing ν. Since Eq. (6) is used for the relaxation data in Fig. 2(d), and the KZ v^{2/3} is derived from τ(K) (Eq. B1–B3), both universal exponents inherit this fragility. The tentative t^{-3/2} decay, being explicitly unbacked by a prediction, is less load-bearing. The method itself may be sound, but the headline claim of universal scaling from real-time dynamics is only as strong as Eq. (6).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a multibath extension of the semigroup influence matrix (SGIM) framework, combining temporal MPS compression of each bath with automated superfermion stitching and spatial MPS compression, and applies it to the two-impurity Anderson model (2IAM) with four fermionic baths. The authors compute spectral functions across the Kondo-to-singlet phase transition, quench relaxation dynamics, and Kibble-Zurek (KZ) ramp statistics. The central claims are that (i) the relaxation time diverges as τ ∼ |K−K_c|^{-2}, matching the two-channel Kondo (2CK) exponent ν=2; (ii) the dissipated work in a KZ ramp obeys ⟨W_d⟩ ∼ v^{2/3}; and (iii) the critical quench decay is tentatively consistent with t^{-3/2}. The methodology is presented as a practical impurity solver for multibath real-time dynamics.","tokens_in":16243,"tokens_out":2534,"duration_ms":26589,"significance":"If the claims hold, the multibath SGIM framework is a significant technical advance: it provides a controlled way to simulate real-time dynamics of quantum impurity models with multiple environments, beyond single-bath influence-matrix solvers. The automated superfermion stitching for fermionic anticommutation across compressed environments, the use of weak symmetries, and the demonstration on a genuinely multibath model (2IAM) are valuable contributions. Reproducing the known 2CK critical exponent and KZ scaling from real-time numerics would be a strong benchmark. However, the headline universal-scaling claims rest heavily on the additive-cutoff correction in Eq. (6), which is assumed rather than derived, and on exponent extractions that lack quantitative error analysis. The method may well be sound, but the paper as written does not yet make the load-bearing part of the case fully convincing.","major_comments":[{"comment":"The central ν=2 claim rests on the additive-cutoff assumption τ^{-1}(K) ≈ τ^{-1}(K;T,χ) − c(T,χ), called a 'natural assumption' and validated only by the collapse in Fig. 2(d). Collapse is a necessary but not sufficient test: a K-dependent cutoff contribution, or a multiplicative cutoff effect, could also produce approximate collapse over a finite K-window while biasing the apparent exponent near K_c. Furthermore, K_c ≈ 0.91Γ is used without a systematic determination or uncertainty estimate; the exponent ν is highly sensitive to K_c in a power-law fit. Please provide: (i) a derivation or at least an independent check of the additive form, e.g., by comparing χ→∞ and T→0 limits separately; (ii) fits using alternative cutoff parameterizations (multiplicative, K-dependent additive) to show the extracted ν is robust; and (iii) error bars on ν propagated from K_c and fitting-range choices.","section":"§4, Eq. (6), Fig. 2(d)"},{"comment":"The exponent extraction from the relaxation-time data is not quantitatively presented. The text states τ diverges as |K−K_c|^{-ν} with ν=2, but Fig. 2(d) shows no fit range, no confidence intervals, and no comparison with neighboring exponents. Given that the collapse in Eq. (6) is the only support for the zero-temperature τ(K), the paper should provide a quantitative fitting procedure, including the K-window used and a χ² or similar measure of the power-law quality. Without this, 'we obtain the universal two-channel-Kondo exponent' is an overstatement of the numerical evidence.","section":"Fig. 2(d), §5"},{"comment":"The t^{-3/2} critical decay is presented as 'consistent with a power law' with the honest caveat that there is no established prediction for this exponent. However, no fit range, error bars, or power-law index estimate are given, and the text does not discuss how the power law is distinguished from a slow exponential or a crossover. Since the abstract and conclusion highlight universal scaling, this tentative claim should either be quantified and clearly labeled as an observation, or removed from the main-text emphasis.","section":"§5, Fig. 2(f)"},{"comment":"The KZ derivation assumes an antisymmetric equilibrium curve ⟨S1·S2⟩_eq around K_c and a scaling ansatz Eq. (B2). While these are plausible, the antisymmetry is not demonstrated for the 2IAM, and the final v^{2/3} scaling inherits the reliability of τ(K) from Eq. (6). Please provide a check of the antisymmetry assumption from the computed equilibrium curves, and discuss how a violation would affect the freeze-out-window integration. This is a secondary issue relative to Eq. (6), but it is part of the load-bearing chain for the KZ claim.","section":"Appendix B, Eqs. (B1)-(B3)"}],"minor_comments":[{"comment":"The spin-1 Kondo temperature comparison in Fig. 2(d) relies on an effective bandwidth D̃ that is chosen to match the data. Since D̃ is a free parameter, the agreement with T_K does not provide an independent validation. Please state this explicitly and, if possible, estimate D̃ from the model parameters.","section":"Appendix C, Eq. (C2)"},{"comment":"The paper does not state data availability or code availability. Given that the method involves a nontrivial automated stitching procedure and numerical implementation, a repository or pseudocode would improve reproducibility.","section":"General"},{"comment":"The caption says 'computed up to t=100/Γ but plotted up to 15/Γ to visibly resolve the decay.' It would be useful to show the full range in a supplementary figure, since the asymptotic relaxation is the quantity used to extract τ.","section":"Fig. 2(e) caption"},{"comment":"Ref. [99] is cited for the additive-rate assumption in Eq. (6). If that reference provides a derivation or justification, it should be explicitly described in the text; otherwise the 'natural assumption' should be flagged as an approximation in need of validation.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper describes a genuinely useful method and the authors have made a serious effort to benchmark it on a nontrivial model. My concerns are not about the method's potential, but about the strength of the evidence for the central universal-scaling claims. The additive-cutoff assumption in Eq. (6) is the linchpin; if it fails, the ν=2 and v^{2/3} results could be numerical artifacts. The authors should be asked to provide a more rigorous validation of Eq. (6) and quantitative error analysis for the exponents. This is within the scope of a major revision, not grounds for rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know: this is a real methodological advance — automated superfermion stitching of multiple SGIMs plus spatial MPS compression, demonstrated on the four-bath two-impurity Anderson model — and the numerics reproduce the expected 2CK scalings. But the central exponent ν=2 claim leans on Eq. (6), an additive-cutoff subtraction the authors call 'natural' and validate only by collapse. That is the load-bearing joint, and it is not derived.\n\nWhat is actually new: Ref. [70] gave the single-bath SGIM. Here they show how to combine independently compressed baths into one dynamical map while preserving fermionic anticommutation, via the superfermion construction and JW reordering with MPO parity strings. That is a non-obvious piece of machinery, and the paper is honest about the technical choices. The spectral functions across the Kondo/2CK/singlet phases and the quench dynamics look like a genuine demonstration. The KZ dissipated-work scaling v^{2/3} following from τ(K) is a nice cross-check, though it inherits the same fragility as Eq. (6).\n\nWhere I would push: the additive-cutoff assumption in Eq. (6) is exactly that — assumed. Collapse of τ(K;T,χ) after subtracting τ^{-1}(Kc;T,χ) is evidence, but not proof: another K-dependence of the cutoff could produce a similar collapse over a finite window with a different apparent exponent. The paper does not provide a systematic Kc determination or error bars on the fitted exponents. The t^{-3/2} decay in Fig. 2(f) is flagged as having no known prediction, so it is more of a tentative observation; that is fine but should not be load-bearing. The effective bandwidth D̃ in App. C is hand-tuned, which weakens the Kondo-temperature comparison. No code is released, which matters for a methods paper.\n\nNone of this sinks the core claim — the method looks real and the collapse across several temperatures and bond dimensions is a meaningful internal check. But the headline 'universal scaling from real-time dynamics' is only as strong as Eq. (6). A referee should ask the authors to either derive the additive form to leading order, or show evidence against plausible alternatives (e.g., a multiplicative or K-dependent cutoff), and to give error estimates on ν.\n\nWho it is for: people working on impurity solvers, real-time DMFT, and non-Markovian quantum dynamics. It deserves a serious referee, with the expectation of a revision that tightens the cutoff analysis.\n\nRecommendation: send to peer review; condition acceptance on addressing the Eq. (6) concern.","headline":"Solid new multibath SGIM stitching, but the headline 2CK exponent rests on an assumed additive-cutoff subtraction that is plausible yet unproven.","tokens_in":16698,"tokens_out":2041,"would_cite":true,"duration_ms":19302,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A tensor-network method for multi-bath impurity dynamics extracts the two-channel Kondo exponent from real-time quenches and ramps.","keywords":["multibath influence matrices","semigroup influence matrix","two-impurity Anderson model","two-channel Kondo critical point","Kibble-Zurek scaling","real-time dynamics","tensor networks","quantum impurity models"],"falsifier":"Compute the inverse relaxation time τ^{-1}(K;T,χ) at fixed low temperature while increasing the SGIM bond dimension χ; if the difference τ^{-1}(K;T,χ) − τ^{-1}(K_c;T,χ) does not converge to a K-dependent function independent of χ, the additive-cutoff assumption fails. Independently, an equilibrium or analytic calculation yielding a critical exponent ν ≠ 2 for the two-channel Kondo point would refute the central claim.","tokens_in":15720,"feed_emoji":"⚛️","tokens_out":8183,"duration_ms":64688,"temperature":0.7,"pith_summary":"The paper introduces multibath semigroup influence matrices, a tensor-network method that simulates real-time dynamics of quantum impurity models coupled to multiple independent fermionic baths. Applied to the two-impurity Anderson model, the method resolves the spectral function across the Kondo-to-singlet transition and tracks quench and ramp dynamics to long times. The central results are a relaxation time diverging as τ ∼ |K−K_c|^−2 at the critical interimpurity exchange and a Kibble-Zurek dissipated work scaling as v^{2/3}, both matching the two-channel Kondo universality class. The authors claim these results establish the method as a practical tool for strongly correlated multi-orbital systems.","feed_headline":"Tensor network extracts two-channel Kondo exponent from real-time dynamics","feed_subtitle":"A multibath influence-matrix method resolves the two-impurity Anderson model's critical point and yields universal scalings.","key_machinery":"The semigroup influence matrix (SGIM) is a single repeated tensor representing a time-translation-invariant environment's entire non-Markovian back-action on an impurity, obtained by infinite-MPS temporal compression. Multiple SGIMs are combined into one dynamical map via an automated 'superfermion stitching' procedure that enforces fermionic anticommutation through parity MPOs; the joint impurity-environment state is further compressed as a spatial matrix product state. The central analytic device is the additive cutoff correction τ^{-1}(K) ≈ τ^{-1}(K;T,χ) − c(T,χ), with c evaluated at K_c, which isolates the genuine zero-temperature relaxation rate from finite-temperature and finite-bond-d","core_discovery":"The paper claims that the nonequilibrium dynamics of the two-impurity Anderson model near its quantum critical point exhibit universal two-channel Kondo scaling, and that these signatures are numerically accessible with the proposed multibath semigroup influence matrix method. Specifically, the quench relaxation time follows τ ∼ |K−K_c|^{-2}, and the dissipated work in a Kibble-Zurek ramp obeys ⟨Wd⟩ ∼ v^{2/3}. Both scalings are read off after a single-parameter collapse in temperature and bond dimension, which extrapolates the finite-cutoff relaxation rates to the zero-temperature line. The paper also reports a possible algebraic critical relaxation ∼ t^{−3/2}, for which no theoretical predi","pith_inferences":["If the additive-cutoff assumption generalizes, the same collapse prescription could extract zero-temperature scaling in other critical open systems where finite-size/temperature cutoffs mask divergences.","The tentative t^{-3/2} critical relaxation, if confirmed by other methods, would constitute a new universal prediction for the 2CK fixed point that field-theoretic calculations have not yet produced.","The Kibble-Zurek dissipated-work protocol maps directly onto charge-Kondo circuits, where the v^{2/3} law could be measured in mesoscopic devices, providing an experimental test of both the method and the 2CK universality class.","The method's independence from the specific number and arrangement of baths opens the door to real-time multi-orbital DMFT simulations of correlated materials, where each orbital is a bath channel."],"forward_implications":["The method resolves transient through asymptotic dynamics across a quantum critical point in a four-bath impurity model, a regime previously beyond reach of tensor-network impurity solvers.","The temperature/bond-dimension collapse validates an additive correction for cutoffs, so zero-temperature critical exponents can be extracted from finite-χ, finite-T simulations of other impurity models.","The observed ν=2 relaxation divergence and v^{2/3} work scaling identify the 2IAM critical point as two-channel Kondo through purely real-time dynamical probes.","Because stitching works for arbitrary bath geometries, the framework extends to chains and tree-tensor impurity setups, and to dissipative or Floquet baths.","The spectral functions computed across the transition provide a benchmark for non-Fermi-liquid signatures that cold-atom and mesoscopic experiments could test."],"fun_headline_variants":["Multibath method catches universal two-channel Kondo scaling","Real-time tensor network maps Kondo critical point","Universal Kondo exponents from quench dynamics","Influence matrices resolve two-impurity Anderson criticality"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is the paper's 'natural assumption' in Eq. (6) that finite temperature and bond dimension add a single K-independent constant to the inverse relaxation time, so subtracting the rate at the critical point removes all cutoffs; this is not derived, and if the cutoffs combine non-additively or K_c is misidentified, the τ ∼ |K−K_c|^{−2} divergence and the 2CK exponent would be artifacts.","fun_headline_variants_meta":{"raw":{"variants":["Multibath method catches universal two-channel Kondo scaling","Real-time tensor network maps Kondo critical point","Universal Kondo exponents from quench dynamics","Influence matrices resolve two-impurity Anderson criticality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000164,"raw_usage":{"total_tokens":1042,"prompt_tokens":665,"completion_tokens":377,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":409,"completion_tokens_details":{"reasoning_tokens":327}},"tokens_in":409,"tokens_out":377,"duration_ms":3891,"temperature":1.0,"reasoning_tokens":327,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T21:00:39.274454+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the inverse relaxation time τ^{-1}(K;T,χ) at fixed low temperature while increasing the SGIM bond dimension χ; if the difference τ^{-1}(K;T,χ) − τ^{-1}(K_c;T,χ) does not converge to a K-dependent function independent of χ, the additive-cutoff assumption fails. Independently, an equilibrium or analytic calculation yielding a critical exponent ν ≠ 2 for the two-channel Kondo point would refute the central claim.","supporting_citations":[],"review_version":1}