{"id":"b8d4d900-8cb7-440b-91f3-473809b377aa","arxiv_id":"2606.19206","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":1,"one_line_summary":"Non-equilibrium quenched AIM dynamics are mapped to an effective non-interacting model with auxiliary static degrees of freedom via numerical optimization on ED and DMRG data.","lead":"The paper maps the time-dependent dynamics of the interacting Anderson Impurity Model after a quench onto an effective non-interacting Gaussian model by adding static auxiliary degrees of freedom. A smart generalist might read it to see whether complex correlated electron problems can be reduced to simpler free-fermion calculations.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Optimization of finite static auxiliaries may fail to reproduce long-time or out-of-sample observables beyond fitting window","rationale":"The reader's weakest_assumption correctly isolates the numerical existence and sufficiency of a finite static auxiliary set. With the full manuscript the same point remains load-bearing because the method is entirely optimization-based; no analytic proof or convergence theorem is invoked, so the practical utility hinges on whether the fit generalizes. This moves the verdict from UNVERDICTED to CONDITIONAL pending the proposed long-time check.","tokens_in":1688,"tokens_out":400,"duration_ms":16900,"concrete_test":"Take the optimized auxiliary parameters obtained from the cost function on [0, t_fit] and recompute the impurity Green function or local moment for t > 2 t_fit using the effective non-interacting Hamiltonian; compare against independent DMRG data at the same longer times. If the L2 deviation exceeds the tolerance achieved inside the fitting window by more than a factor of two, the static-auxiliary mapping is limited to short-time reproduction.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that a finite collection of time-independent auxiliary fermionic modes, with numerically optimized hoppings and on-site energies, can be coupled to the non-interacting AIM such that the resulting Gaussian dynamics match the interacting quench evolution (obtained via ED/DMRG) to useful accuracy. Because the auxiliaries are constrained to be static, any time-dependent effective interactions generated by the original U term must be absorbed into the enlarged single-particle spectrum and initial state. The numerical optimization procedure therefore implicitly assumes that the required auxiliary Hilbert space dimension remains modest and that the fit does not degrade outside the time interval or parameter set used for the cost function; neither an a-priori bound on the required dimension nor a demonstration that the mapping remains accurate for t ≫ t_fit or for different quench protocols is supplied by the construction itself.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper claims that the time-dependent dynamics of the interacting Anderson Impurity Model (AIM) following a quench can be reproduced by an effective non-interacting Gaussian theory obtained by coupling the original impurity and bath to a finite set of additional static auxiliary fermionic modes whose hoppings and on-site energies are determined by numerical optimization against ED and DMRG data.","tokens_in":1884,"tokens_out":501,"duration_ms":16202,"significance":"If the mapping can be shown to require only a modest number of auxiliaries and to remain accurate outside the fitting window, it would provide a practical route to interpret and compute non-equilibrium impurity dynamics via solvable Gaussian methods while revealing structure in the auxiliary spectrum. The direct numerical matching to exact solvers is a concrete strength of the approach.","major_comments":[{"comment":"The optimization procedure (described in the methods and results sections) supplies no a-priori bound on the required auxiliary Hilbert-space dimension, no convergence criteria, and no error bars on the fitted parameters. Because the central claim rests on the existence of a finite static auxiliary set that reproduces the interacting quench evolution, these details are load-bearing for assessing whether the mapping is robust or merely a post-hoc fit.","section":"Methods and optimization procedure"},{"comment":"No tests are reported for times t ≫ t_fit or for quench protocols different from those used in the cost function. The skeptic concern that static auxiliaries may fail to capture long-time or out-of-sample observables therefore remains unaddressed, directly affecting the generality of the claimed mapping.","section":"Results on time evolution"},{"comment":"The manuscript does not quantify how the quality of the fit (e.g., deviation in local observables or spectral functions) scales with the number of auxiliary modes, leaving open whether the effective dimension remains modest for physically relevant parameters.","section":"Numerical results"}],"minor_comments":[{"comment":"Clarify the precise form of the effective single-particle Hamiltonian, including how the auxiliary modes are coupled to the original AIM degrees of freedom.","section":"Model definition"},{"comment":"Add a table or figure summarizing the number of auxiliaries used for each parameter set and the achieved fit accuracy.","section":"Results"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and constructive comments on our manuscript. We address each major comment below and indicate the revisions we will make.","responses":[{"response":"We agree that the optimization procedure requires fuller documentation. In the revised manuscript we will expand the methods section to specify the explicit form of the cost function, the numerical optimizer used, the convergence tolerance applied to the cost, and error estimates on the fitted parameters obtained from repeated optimizations with varied initial conditions. An analytic a-priori bound on auxiliary dimension is not available, but we will add a discussion of the empirical criterion (saturation of fit quality with increasing auxiliary number) already used to select the reported dimensions.","revision_made":"yes","referee_comment":"[Methods and optimization procedure] The optimization procedure (described in the methods and results sections) supplies no a-priori bound on the required auxiliary Hilbert-space dimension, no convergence criteria, and no error bars on the fitted parameters. Because the central claim rests on the existence of a finite static auxiliary set that reproduces the interacting quench evolution, these details are load-bearing for assessing whether the mapping is robust or merely a post-hoc fit."},{"response":"The referee correctly notes the absence of out-of-sample tests. We will add new calculations that propagate the fitted auxiliary model to times t ≫ t_fit and to quench protocols not included in the original cost function, comparing the resulting observables against available ED or DMRG reference data. These results will be included in a new subsection of the results.","revision_made":"yes","referee_comment":"[Results on time evolution] No tests are reported for times t ≫ t_fit or for quench protocols different from those used in the cost function. The skeptic concern that static auxiliaries may fail to capture long-time or out-of-sample observables therefore remains unaddressed, directly affecting the generality of the claimed mapping."},{"response":"We will add a systematic quantification of fit quality versus auxiliary number. A new figure will show the scaling of the mean-squared deviation in local observables (and, where computable, spectral functions) as a function of auxiliary count for several values of interaction strength and bath parameters, thereby demonstrating that the required dimension remains modest in the regimes studied.","revision_made":"yes","referee_comment":"[Numerical results] The manuscript does not quantify how the quality of the fit (e.g., deviation in local observables or spectral functions) scales with the number of auxiliary modes, leaving open whether the effective dimension remains modest for physically relevant parameters."}],"tokens_in":1346,"tokens_out":557,"duration_ms":28160,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main takeaway is that the authors take existing ED and DMRG solutions for the quenched Anderson impurity model and tune the hoppings and on-site energies of a handful of extra static fermionic modes until the single-particle dynamics of the enlarged Gaussian system reproduce the interacting observables.\n\nWhat is new is the application to the non-equilibrium quench: they start from the full interacting time evolution and back out an effective non-interacting representation with time-independent auxiliaries. The paper reports that the optimization produces some recognizable structure in the auxiliary parameters, which is at least an interesting observation even if it is post-hoc.\n\nThe approach works for the cases they examine because the fit is performed directly against the numerical data. That is the strength: once the auxiliaries are fixed, any observable that can be computed in the non-interacting model becomes available without further interacting calculations.\n\nThe soft spot is exactly the one flagged in the stress test. Because the auxiliaries are constrained to be static, any effective time dependence generated by the original interaction must be absorbed into the enlarged single-particle spectrum and initial state. The construction therefore depends on the numerical optimization succeeding, yet the abstract and the description give no information on the number of auxiliaries required, the time window used for the cost function, error bars as a function of time, or tests on quenches outside the training set. If the match degrades for t much larger than the fitting interval or for different interaction strengths, the claim that the dynamics “can be described” by the Gaussian model is limited to the fitted regime.\n\nThis is for people already working on non-equilibrium impurity problems who want an alternative computational or interpretive route. A reader who needs a first-principles mapping or guaranteed long-time accuracy will not find it here.\n\nI would send it to peer review. The numerical evidence is concrete enough to be worth referee scrutiny on the robustness of the fits and the scope of the mapping.","headline":"They numerically optimize a finite set of static auxiliary fermions so the enlarged non-interacting AIM matches ED/DMRG quench data, but the mapping has no analytic basis and its accuracy outside the fit window is unproven.","tokens_in":2404,"tokens_out":477,"would_cite":false,"duration_ms":17048,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"The quench dynamics of the interacting Anderson Impurity Model can be reproduced exactly by a non-interacting Gaussian model coupled to static auxiliary degrees of freedom.","keywords":["Anderson Impurity Model","non-equilibrium quench","Gaussian mapping","auxiliary degrees of freedom","numerical optimization","effective non-interacting theory","ED DMRG comparison"],"falsifier":"Demonstration that, for any finite number of auxiliary sites, no choice of couplings reproduces the time-dependent local magnetization or occupation number from the full interacting model within a target error would falsify the mapping.","tokens_in":2574,"feed_emoji":"⚛","tokens_out":733,"duration_ms":20611,"temperature":0.7,"pith_summary":"The paper establishes that the time-dependent evolution of the Anderson Impurity Model following a quench, which normally requires handling strong electron correlations, can instead be captured by an enlarged non-interacting model. The extra static auxiliary sites are introduced and their couplings are chosen through numerical optimization so that observables match those obtained from exact diagonalization and DMRG on the original interacting system. A sympathetic reader would care because this converts a difficult interacting non-equilibrium problem into one that can be solved with standard free-fermion techniques on a larger but Gaussian Hamiltonian. The approach uncovers structure within the auxiliary system and offers a route to compute quench properties without directly simulating the interactions.","feed_headline":"Quench dynamics of interacting AIM reduced to non-interacting Gaussian model","feed_subtitle":"Static auxiliary sites with optimized couplings reproduce the full interacting evolution after a quench.","key_machinery":"The effective non-interacting Anderson Impurity Model augmented by a finite set of static auxiliary degrees of freedom whose couplings are numerically optimized to reproduce the interacting quench dynamics.","core_discovery":"The time-evolving dynamics of the AIM after a quench can be described by a completely non-interacting version of the model, at the expense of coupling to additional static auxiliary degrees of freedom. Starting from the full solution of the quenched AIM using ED and DMRG, the properties of this mapping are studied using numerical optimization, and intriguing structure in the auxiliary system is uncovered. The method allows understanding interacting non-equilibrium dynamics through the simpler lens of an effective non-interacting system of larger dimension.","pith_inferences":["The auxiliary sites may serve as an effective discrete representation of the bath that could be reused for other quench protocols or steady-state calculations.","The approach could be tested on longer evolution times or different interaction strengths by comparing optimized results against larger-scale DMRG runs.","If the auxiliary structure is universal, the mapping might extend to other impurity models or to dynamical mean-field theory embeddings without re-deriving the optimization each time.","Hybrid solvers could combine this Gaussian mapping with perturbative corrections for regimes where the auxiliary count must remain small."],"forward_implications":["Quench dynamics and observables can be computed exactly using free-fermion methods on the enlarged Gaussian system.","The mapping reproduces key time-dependent quantities to useful accuracy once the auxiliary couplings are optimized.","Structure within the auxiliary degrees of freedom becomes visible and can be characterized directly.","The same non-interacting enlargement applies to any observable that can be obtained from the original interacting solution."],"fun_headline_variants":["Quench AIM dynamics mapped to non-interacting Gaussian with auxiliaries","AIM quench recast as larger non-interacting model via static auxiliaries","Non-equilibrium AIM mapped to Gaussian theory using optimized auxiliaries","Quenched interacting AIM reduced to effective non-interacting system"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"A finite collection of static auxiliary degrees of freedom exists whose couplings can be chosen by numerical optimization to match the interacting quench dynamics to useful accuracy on the observables of interest.","fun_headline_variants_meta":{"raw":{"variants":["Quench AIM dynamics mapped to non-interacting Gaussian with auxiliaries","AIM quench recast as larger non-interacting model via static auxiliaries","Non-equilibrium AIM mapped to Gaussian theory using optimized auxiliaries","Quenched interacting AIM reduced to effective non-interacting system"]},"model":"grok-4.3","cost_usd":0.00389,"raw_usage":{"total_tokens":1901,"prompt_tokens":636,"num_sources_used":0,"completion_tokens":70,"cost_in_usd_ticks":38903000,"prompt_tokens_details":{"text_tokens":636,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1195,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":636,"tokens_out":70,"duration_ms":6098,"temperature":1.0,"reasoning_tokens":1195,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-26T18:55:46.625279+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Demonstration that, for any finite number of auxiliary sites, no choice of couplings reproduces the time-dependent local magnetization or occupation number from the full interacting model within a target error would falsify the mapping.","supporting_citations":[],"review_version":1}