{"id":"f442595d-2fb2-475b-87cc-d85aab56d293","arxiv_id":"2501.06752","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":1.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A community roadmap reviewing the state of theoretical and computational modeling of ultrafast phenomena in quantum materials, with no new research results.","lead":"This paper is a multi-author roadmap review of theoretical and computational methods for ultrafast dynamics in quantum materials. It surveys ab initio spectroscopy, Floquet engineering, cavity QED, nonequilibrium Green's functions, and FAIR data efforts.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3's linear-scaling G1-G2 claim stands only for single-time observables; the roadmap itself concedes the loss of correlated spectral functions, and no proof exists that spectral restoration will keep the scaling advantage.","rationale":"The reader's weakest_assumption correctly identifies that the G1-G2 scheme loses access to correlated spectral functions, and that the roadmap's own text lists spectral restoration as a major uncompleted task. This is the same load-bearing concern I find: the central claim of Section 3 is technically accurate for time-diagonal quantities, but the roadmap's advertised practical advantage—combining NEGF's descriptive power with TDDFT's cost—depends on restoring spectra without sacrificing linear scaling, which is neither proven nor analyzed. The roadman's internal caveat is explicit, so there is no hidden inconsistency; however, the concern is real because multiple sections of the roadmap (Sections 2, 8, 9) rely on spectral functions as the key experimental link. My concrete test would settle whether spectral restoration can preserve the linear-scaling advantage. Since the roadmap is an invited perspective rather than a research claim, the appropriate verdict remains UNVERDICTED, and I do not propose changing the reader's verdict.","tokens_in":66357,"tokens_out":5418,"duration_ms":57927,"concrete_test":"On a 4-site Hubbard model, compute the spectral function A(k, ω) from (i) the G1-G2 time-local scheme alone, (ii) the G1-G2 scheme augmented by the time-off-diagonal propagation of Ref. [23]. Measure CPU time versus N_t from 10^2 to 10^5 steps, fit the observed scaling exponent, and compare the resulting spectra to an exact diagonalization benchmark. If restoring spectra changes the scaling from ~N_t to ~N_t^2 or higher, then the linear-scaling claim is restricted to single-time observables and does not support the roadmap's statement that the scheme 'matches' TDDFT for the full NEGF program.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most concrete technical claim in this roadmap, Section 3, is that the G1-G2 scheme 'exactly reformulates the HF-GKBA in time-local form,' achieves 'linear scaling in N_t,' and 'has since changed the field of NEGF simulations.' The load-bearing assumption for any practical impact is that the linear-scaling speedup extends to the observables that make NEGF valuable, especially correlated spectral functions measured by TR-ARPES (Sections 2, 8, 9). Yet the same section explicitly states: 'The linear scaling approach (as the standard HF-GKBA) loses access to high-quality correlated spectral functions which is a major advantage of the NEGF approach [15].' The proposed fixes, such as Koopmans' theorem or 'time-off-diagonal propagation' (Ref. [23]), are listed in 'Current and Future Challenges' as requiring implementation and testing, not as proven extensions. In particular, restoring the two-time Green's function via time-off-diagonal propagation would reintroduce the two-time memory cost that the G1-G2 scheme was designed to eliminate; if that cost scales as N_t^2 or worse, the claimed parity with time-dependent DFT no longer holds for spectral properties. The roadmap offers no analysis of this trade-off, so the central efficiency claim is conditional on an unverified restoration mechanism.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a collective roadmap comprising 14 contributed sections on theoretical and computational modelling of ultrafast dynamics in quantum materials. The contributions cover nonequilibrium Green's function methods (sections 2-5), Floquet and cavity-based materials engineering (sections 6-8), ab initio time-resolved spectroscopy and magnetization dynamics (sections 9-10), phonon dynamics (sections 11-12), and FAIR data protocols for time-resolved experiments and simulations (sections 13-14). Each section follows a common structure of current status, open challenges, and possible advances, with the overall goal of identifying key problems whose solution would advance the field.","tokens_in":66618,"tokens_out":5431,"duration_ms":52744,"significance":"As a roadmap, the paper's value lies in the expertise of its contributors and in its explicit, useful cataloguing of open problems and methodological bottlenecks. Several sections are commendably candid: Section 3 acknowledges that the linear-scaling G1-G2 scheme loses access to high-quality correlated spectral functions, and Section 4 highlights the lack of numerically exact nonequilibrium impurity solvers. The breadth of coverage and the consistent section structure make it a convenient entry point for researchers seeking the state of the art in this interdisciplinary area. The manuscript does not present new derivations or data, but this is appropriate for a roadmap rather than a research article.","major_comments":[{"comment":"The statement that the G1-G2 scheme achieves computational efficiency matching time-dependent DFT is unqualified and sits in tension with the same section's earlier admission that the linear-scaling approach loses access to high-quality correlated spectral functions, which is described as a major advantage of the NEGF approach. Because those spectral functions are central to the time-resolved spectroscopies emphasized in Sections 8 and 9, the parity claim should be restricted to single-time observables and should note that spectral reconstruction methods, such as time-off-diagonal propagation, have not yet been shown to preserve the linear scaling.","section":"Section 3, Concluding Remarks"},{"comment":"The proposed route of including time-off-diagonal propagation to restore spectral information is listed without any discussion of its computational cost. Since reintroducing the two-time Green's function is precisely what the G1-G2 scheme avoids, the roadmap should either explain why this restoration route might retain the linear-scaling advantage or explicitly identify the scaling of spectral restoration as an open question; as written, the reader cannot assess whether the central efficiency claim extends to any observable beyond the single-time density matrix.","section":"Section 3, Advances in Science and Technology to Meet Challenges"}],"minor_comments":[{"comment":"In the final paragraph, 'so as to threat more realistic systems' should read 'so as to treat more realistic systems'.","section":"Section 9, Concluding Remarks"},{"comment":"Reference [8] lists the DOI 10.1103/PhysRevB.101.245101, but the cited paper (Schlünzen, Joost, and Bonitz, Phys. Rev. Lett. 124, 076601, 2020) has DOI 10.1103/PhysRevLett.124.076601; the DOI currently given belongs to reference [9].","section":"Section 3, References"},{"comment":"Reference [14] contains the malformed DOI string '10.1103/PhysRevLett.127.036402-036402-(2021)' and should be corrected to '10.1103/PhysRevLett.127.036402'.","section":"Section 3, References"},{"comment":"The citation key [Sch22] is used for two distinct references (Schlawin et al. and Scheffler et al.); this colliding key should be disambiguated to avoid confusion in the reference list.","section":"Section 1, References"},{"comment":"The caption reads 'upon impact or a highly charged ion' and should read 'upon impact of a highly charged ion'.","section":"Section 3, Figure 2 caption"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is a roadmap rather than a research article, so I have assessed it for balance, accuracy of the literature snapshot, and usefulness of the identified challenges. The self-citation concentration in Section 3 is noticeable but is a natural consequence of that section being authored by the developers of the G1-G2 method; importantly, the section also states the method's limitations. My recommendation of minor revision is driven by the need to qualify one overbroad efficiency claim and to fix a small number of typographical and reference errors."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a roadmap, not a research paper, so read it as an expert inventory plus agenda-setting exercise. On that metric it works: fourteen short sections from the people actually building these methods, covering NEGF, DMFT, Floquet, QEDFT, tensor networks, and FAIR data, each with a status statement and a challenges list. The structure is genuinely useful for orientation, and the field is lucky to have this many leading groups write down where things stand.\n\nWhat is genuinely good: the document is mostly transparent about its own limitations. Section 3, for instance, advertises the G1-G2 linear-scaling scheme, but it explicitly states that the approach loses access to high-quality correlated spectral functions, which is a major advantage of NEGF. The stress-test is right that restoring those spectral functions via time-off-diagonal propagation risks bringing back the two-time cost the method was designed to eliminate, and the roadmap does not analyze that trade-off. But that is a future challenge the authors themselves flag, not a hidden assumption. My main complaint is rhetorical: saying G1-G2 'has since changed the field' overstates what a scheme limited to single-time observables has actually delivered, and the Status section would be stronger with that caveat in the same paragraph instead of buried in Current and Future Challenges.\n\nThe soft spots are what you would expect from a self-authored roadmap. Heavy self-citation is inherent to the format, since each section is written by the leading contributor to that subfield, and it shows in some evaluative phrases. There is no independent benchmarking of competing methods and no attempt to adjudicate between NEGF, DMFT, and tensor network approaches. If you want a technical result, go to the cited papers; this document only points you there. It also does not claim any new derivations or data, so novelty is intentionally zero.\n\nWho gets value: a graduate student entering ultrafast theory, an experimentalist wanting a map of what ab initio methods can and cannot do, a funding panel looking for consensus on open problems. It deserves a serious referee, because a roadmap that cites and names the state of the field should be checked for accuracy and fair representation. My recommendation: accept after minor revision, mainly to temper 'changed the field' to something like 'substantially reduced the cost for single-time observables' and to add the spectral-function caveat next to the scaling claim. This is a useful reference document, not a breakthrough, and it should be treated as such.","headline":"A credible, useful field inventory that is honest about most of its limits, though the G1-G2 'changed the field' claim is stronger than its own spectral-function caveat warrants.","tokens_in":67235,"tokens_out":1916,"would_cite":false,"duration_ms":24849,"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":"The G1-G2 scheme makes correlated electron dynamics simulations scale linearly in simulation time.","keywords":["ultrafast dynamics","nonequilibrium Green's functions","G1-G2 scheme","linear-scaling algorithms","strongly correlated systems","Floquet engineering","cavity materials engineering","time-resolved spectroscopy"],"falsifier":"A direct falsifier: implement G1-G2 with time-off-diagonal propagation on a small Hubbard chain or two-dimensional cluster where exact diagonalization provides the exact spectral function; if the restored spectral functions deviate systematically, or the cost grows faster than linearly when spectral accuracy is required, the practical value of the linear-scaling claim is refuted.","tokens_in":66194,"feed_emoji":"⚡","tokens_out":7311,"duration_ms":66085,"temperature":0.7,"pith_summary":"This Roadmap sets out the current frontier of theoretical and computational modelling of ultrafast phenomena in quantum materials, and its most concrete internal claim is about algorithm speed. Section 3 argues that the G1-G2 scheme exactly reformulates the Hartree-Fock generalized Kadanoff-Baym ansatz (HF-GKBA) in time-local form, cutting the CPU cost of nonequilibrium Green's function (NEGF) simulations from cubic to linear scaling in the number of time steps $N_t$, and enabling the same linear scaling for GW and T-matrix self-energies. If correct, correlated electron dynamics can be simulated for durations that were previously out of reach, at a cost comparable to time-dependent density functional theory, which would matter for predicting long-time relaxation, light-induced phases, and transient spectra. The price, stated openly, is that the time-local form loses high-quality correlated spectral functions, and the scheme has large memory requirements from the two-particle quantity G2.","feed_headline":"Correlated electron dynamics now scale linearly in simulation time","feed_subtitle":"A time-local reformulation of nonequilibrium Green's functions unlocks long ultrafast runs","key_machinery":"The G1-G2 scheme is a time-local, linear-scaling reformulation of the HF-GKBA for nonequilibrium Green's functions. The one-particle Green's function G1 is propagated together with the correlated part of the two-particle Green's function G2, which acts as the source of correlation; working entirely on the time diagonal removes the Kadanoff-Baym memory integrals that made standard NEGF cost scale cubically in the simulation duration. This scheme is the central object because it is what turns the linear-scaling claim from a formal possibility into an implemented method for GW, T-matrix, and dynamically screened ladder self-energies.","core_discovery":"The central claim is the G1-G2 scheme: two coupled, time-local equations of motion, one for the one-particle Green's function G1 and one for the correlated part of the two-particle Green's function G2, that exactly reproduce the HF-GKBA without storing or integrating two-time products over the whole history. Because the memory integrals vanish, CPU time scales linearly with the simulation duration $N_t$, and the same construction is shown to work for GW, T-matrix, and their self-consistent dynamically screened ladder combination. The scheme's stated limitation is that it loses access to high-quality correlated spectral functions, inherited from HF-GKBA, and that G2 as a rank-4 tensor raises memory and basis-scaling costs; the roadmap presents embedding, tensor compression, and the quantum fluctuations approach as the proposed remedies. On the authors' terms, the consequence is a NEGF method whose efficiency matches time-dependent DFT while retaining a diagrammatic treatment of correlations.","pith_inferences":["If spectral restoration succeeds without breaking linear scaling, NEGF simulations could become cheap enough that many-body corrections become a standard part of ultrafast materials screening rather than a specialized calculation.","The quantum fluctuations approach, by compressing G2, may effectively eliminate the purification overhead and extend G1-G2 from lattice models to first-principles basis sets; this is the roadmap's own expectation, not yet a demonstrated result.","The same linear-scaling idea could accelerate first-principles Floquet and cavity studies, whose bottleneck is also long-time correlated dynamics; that connection is not developed in the road map beyond cross-references.","A direct benchmark on small Hubbard clusters comparing G1-G2 spectral functions restored by time-off-diagonal propagation against exact diagonalization would settle whether the spectral-information loss is practically recoverable."],"forward_implications":["Long-time simulations of laser-driven correlated electrons become feasible at linear cost, so relaxation, prethermalization, and slow collective dynamics are no longer excluded by cubic scaling.","Linear scaling extends beyond the simplest self-energy to GW and T-matrix approximations, and to their self-consistent dynamically screened ladder combination, broadening the class of problems that can be treated.","Direct access to two-particle observables such as pair distributions is a by-product of propagating G2, which is useful for probing correlations during the dynamics.","The practical bottleneck shifts from CPU time to memory and basis-size scaling, with embedding and tensor-compression routes already proposed to reduce the cost of the rank-4 tensor.","A stated near-term priority is restoring spectral information via nonequilibrium Koopmans' theorem, time-off-diagonal propagation, or machine-learned self-energies."],"supporting_citations":[{"why":"Defines the generalized Kadanoff-Baym ansatz that G1-G2 exactly reformulates; without this the time-local equations have no target.","marker":"[6]"},{"why":"Establishes the scaling limit: the G1-G2 scheme cuts CPU time to linear scaling in the simulation duration.","marker":"[8]"},{"why":"Shows the scheme accelerates HF-GKBA simulations and retains linear scaling for GW and T-matrix self-energies.","marker":"[9]"},{"why":"Introduces the dynamically screened ladder approximation within G1-G2, combining GW and particle-particle and particle-hole T-matrix diagrams self-consistently.","marker":"[10]"},{"why":"Recent overview that documents applications, limitations, and the need to generate correlated initial states by adiabatic switching.","marker":"[11]"},{"why":"States the loss of high-quality correlated spectral functions as a limitation inherited from HF-GKBA.","marker":"[15]"},{"why":"Demonstrates an embedding variant that restricts the full-accuracy G2 to a subset of states and preserves time-linear scaling.","marker":"[17]"},{"why":"Introduces the quantum fluctuations approach as a physically motivated compression of G2, avoiding the scaling penalty of purification.","marker":"[20]"},{"why":"Shows the quantum fluctuations approach applied to nonequilibrium GW, giving density correlations and the dynamic structure factor stably.","marker":"[22]"}],"fun_headline_variants":["Memory-free NEGF: linear-time ultrafast dynamics","G1-G2 scheme: NEGF scales linearly in time","Time-local NEGF: linear scaling for long runs","Linear-scaling NEGF matches DFT efficiency","From two-time to time-local: NEGF speedup"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claimed practical advantage of G1-G2 collapses if the lost correlated spectrum cannot be recovered without re-introducing two-time memory costs severe enough to cancel the linear speed-up.","fun_headline_variants_meta":{"raw":{"variants":["Memory-free NEGF: linear-time ultrafast dynamics","G1-G2 scheme: NEGF scales linearly in time","Time-local NEGF: linear scaling for long runs","Linear-scaling NEGF matches DFT efficiency","From two-time to time-local: NEGF speedup"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000842,"raw_usage":{"total_tokens":3661,"prompt_tokens":930,"completion_tokens":2731,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":546,"completion_tokens_details":{"reasoning_tokens":2654}},"tokens_in":546,"tokens_out":2731,"duration_ms":22340,"temperature":1.0,"reasoning_tokens":2654,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:49:08.681301+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct falsifier: implement G1-G2 with time-off-diagonal propagation on a small Hubbard chain or two-dimensional cluster where exact diagonalization provides the exact spectral function; if the restored spectral functions deviate systematically, or the cost grows faster than linearly when spectral accuracy is required, the practical value of the linear-scaling claim is refuted.","supporting_citations":[],"review_version":1}