{"id":"c113fe2a-7274-4ef1-8b09-be763b8747b2","arxiv_id":"2412.13332","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"WaveguideQED.jl is an efficient, open-source, time-bin simulator for waveguide QED with up to two photons, including full photonic states and non-Markovian feedback.","lead":"This paper introduces WaveguideQED.jl, an open-source Julia package that simulates traveling photons interacting with small quantum systems using a time-bin collision picture. It makes few-photon waveguide quantum electrodynamics, including non-Markovian feedback such as mirrors, practical and fast to set up.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Distinctive two-photon and delayed-feedback capabilities are validated only qualitatively; a single-photon analytic benchmark does not cover the key new features.","rationale":"The reader's weakest assumption concerned the flat spectral density and two-photon limit; those are explicit scope restrictions, not internal flaws. My concern is different: the quantitative evidence is thinnest exactly where the framework departs from standard one-photon Markovian scattering. The single-photon EOM benchmark in Appendix A validates the time-bin discretization, but the two-photon basis and delayed feedback operators are the features that make the package distinctive, and they are only checked qualitatively against expected physics. Because the code is open source and the examples are reproducible, this is a verification gap rather than a demonstrated error. The verdict should stand, but a targeted independent benchmark would materially strengthen the acceptance.","tokens_in":18987,"tokens_out":18870,"duration_ms":205498,"concrete_test":"Reproduce Code Sample 3 with Δt = 0.05 and Δt = 0.025, and compare the converged two-photon output ξ_out(t,t') and its Schmidt weights against an independent MPS or analytic calculation for two-photon scattering by a two-level emitter in a 1D waveguide; then reproduce Code Sample 5 with τ = 1/γ and compare the emitter population to an independent delayed-feedback solver or a real-space discretization at much smaller Δt. Agreement at the level of the Fig. 7 single-photon benchmark would settle the concern.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that WaveguideQED.jl is a general, efficient tool for two-photon waveguide QED with non-Markovian feedback. The only quantitative check in the paper is the single-photon scattering convergence test in Appendix A (Fig. 7), where the model reduces to Eqs. (16)-(17). The framework's distinctive machinery—the symmetrized two-photon basis in Eq. (9), the bosonic normalization of w_k acting on |2_i> and |1_i1_k>, and the delayed operators in Eq. (20) with integer delay tau/dt—is never validated against an independent two-photon or feedback result. An indexing or normalization error in that machinery would produce plausible-looking but quantitatively wrong correlations (Figs. 3-5), directly undermining the 'general' and 'robust' claims. This is an evidence gap, not a demonstrated inconsistency; the open-source code and single-photon benchmark are real support, but they do not exercise the advertised two-photon or non-Markovian sectors beyond self-consistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces WaveguideQED.jl, an open-source Julia framework for simulating waveguide quantum electrodynamics based on the collision model. The traveling field is discretized into time-bins, the waveguide state is represented in a sparse basis restricted to at most two photons, and the waveguide operators are implemented as matrix-free kernel functions combined with LazyOperators from QuantumOptics.jl. The manuscript demonstrates single- and two-photon scattering off a two-level emitter, a two-waveguide splitter setup, delayed-feedback dynamics for an emitter in front of a mirror, and benchmarks against a naive sparse-matrix implementation and a space-discretized model. The only quantitative convergence check, in Appendix A, compares the single-photon scattering result against an analytical equation of motion derived from the same discretized Hamiltonian.","tokens_in":19163,"tokens_out":7629,"duration_ms":69253,"significance":"If the framework performs as claimed, it fills a useful niche: a fast, open, intuitive tool for full two-photon wavefunctions with delayed feedback, backed by complete code samples and a public repository. The design choices are sensible—restricted two-excitation basis, lazy operators, kernel-based operator application—and the single-photon convergence test is a genuine strength. The main caveat is that the two-photon and delayed-feedback sectors, which are the distinctive advertised capabilities, lack independent quantitative validation; correctness in those sectors currently rests on internal consistency and qualitative plausibility. The acknowledged limitations (flat spectral density, two-photon maximum, quasi-constant wavefunction per time-bin) are stated clearly and are appropriate for the intended scope.","major_comments":[{"comment":"The distinctive two-photon and delayed-feedback capabilities are validated only qualitatively. The convergence test in Appendix A (Fig. 7) exercises the single-photon sector, where the framework reduces to Eqs. (16)-(17); the two-photon basis in Eq. (9), the action of the waveguide operators on two-photon states, and the delayed operators in Eq. (20) are never benchmarked against an independent analytic or numerical solution. An indexing or normalization error in this machinery would produce plausible-looking but quantitatively wrong correlations in Figs. 3-5, directly affecting the advertised generality. I request at least one quantitative external check, e.g., comparing the two-photon scattering in Fig. 3 with known analytic two-photon scattering solutions for a two-level emitter, and comparing the feedback dynamics in Fig. 5 with the published results of Ref. [50] or with an independent space-discretized simulation using the same parameters.","section":"Appendix A; Sec. 3.2; Sec. 4.2"},{"comment":"The displayed Hamiltonians in Eqs. (19) and (20) are inconsistent with Eq. (10) and with the corresponding code samples. Eq. (10) gives H_time = sum_k f_k(t) i hbar sqrt(gamma/Delta t)(sigma^dag w_k - sigma w_k^dag), and Code Samples 4 and 5 use `im*sqrt(...)*(...)`. Eqs. (19) and (20) omit the imaginary unit and the hbar factor, and Eq. (19) also uses a plus sign where Eq. (10) has a minus sign. As written, a reader implementing from these equations would obtain different dynamics (Hermitian vs. non-Hermitian, or with the wrong relative phase). The equations should be corrected to match the interaction-picture Hamiltonian used in the code, or the notation should be explicitly defined if a different convention (e.g., hbar=1 and a particular sign convention) is intended.","section":"Eqs. (19)-(20) vs. Eq. (10) and Code Samples 4-5"},{"comment":"The statement that \"the memory usage of WaveguideQED.jl is constant in the number of time-bins N\" is incorrect for the two-photon simulations showcased in the paper. The state vector for a two-photon waveguide contains O(N^2) amplitudes (as given in Sec. 5.3: N(N+1)/2 elements per waveguide mode), so memory necessarily grows with N. The likely intended point is that the operator implementation does not require preallocating N sparse matrices; this should be stated precisely to avoid a false scalability claim.","section":"Sec. 5.2, memory scaling claim"}],"minor_comments":[{"comment":"There are several typos: \"WagveguideQED.jl\" should be \"WaveguideQED.jl\", and \"the two solutions agreement agree very well\" should read \"the two solutions agree very well\".","section":"Sec. 3, first paragraph; Sec. 3.2"},{"comment":"The emitter operators are written as a and a^dag in Eq. (13) without definition, while the rest of the paper uses sigma and sigma^dag; please align the notation or define a explicitly as the emitter lowering operator.","section":"Eq. (13)"},{"comment":"The total element count for multiple waveguides omits the N diagonal two-photon states per waveguide. The count should be N_W * N(N+1)/2 for two photons in the same waveguide plus N_W(N_W-1)/2 * N^2 for two photons in different waveguides; the displayed expression \"N_W N(N-1)/2 + N_W N^2\" misses the N diagonal terms.","section":"Sec. 5.3, element counting"},{"comment":"The term \"non-Markovian\" may mislead readers without qualification: the framework assumes a flat, frequency-independent system-bath coupling, and memory effects enter only through engineered delays. This is stated in Sec. 4.2, but a one-sentence qualifier in the abstract would help set expectations.","section":"Abstract and Sec. 4.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript header indicates acceptance in Quantum; my recommendation concerns the technical state of the present version. The main risk is the absence of independent quantitative validation for the two-photon and delayed-feedback sectors, which are the paper's defining features. If the authors add one external benchmark for each of those sectors and fix the Hamiltonian display errors, I would consider the manuscript acceptable. The open-source repository and reproducible code samples are strong points in favor of publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague——\n\nThis is a solid software-contribution paper. The genuinely new thing is the package itself: an open-source, high-performance realization of the quantum collision model for waveguide QED, built on QuantumOptics.jl, with matrix-free time-dependent operators and a tailored few-excitation basis. The authors show orders-of-magnitude speedups over naive sparse-matrix evolution and a working delayed-feedback capability. If you work in waveguide QED and want to simulate two-photon scattering or mirror feedback without writing your own discretized code, this tool is genuinely useful and the code is available.\n\nThe physics is not new: collision models are established, and the flat-coupling assumption (g(ν)=i√(γ/2π), stated after Eq. 3) means the environment is Markovian at the level of the system–bath coupling; memory enters only through engineered delays. The paper is transparent about that scope in Sec. 5.4, so it is a fair limitation, not a hidden one.\n\nThe soft spot is exactly what the stress-test note identifies: the only quantitative validation is the single-photon convergence check in Appendix A, which confirms the discretization and the operator kernel implementation in the one-photon sector. The distinctive machinery—the symmetrized two-photon basis in Eq. (9), the bosonic normalization of w_k acting on |2_i> and |1_i1_k>, and the delayed operators in Eq. (20) with integer delay τ/dt—is never checked against an independent two-photon or feedback result. The figures look physically plausible (the excitation trapping in Fig. 5c matches known qualitative behavior), but a normalization or indexing bug in the two-photon basis could produce plausible-looking but quantitatively wrong correlations without tripping the single-photon check. That is an evidence gap, not a demonstrated error; the code is open and the authors' track record is good, so I would guess the machinery is correct. Still, a referee should ask for at least one external cross-check—an MPS calculation or an analytic two-photon scattering matrix—before the two-photon outputs are taken as quantitatively reliable.\n\nThe performance comparison with Ref. [14] is a bit rough (different method, different hardware), but it is presented as indicative, and the naive-matrix benchmark is controlled and convincing.\n\nBottom line: this deserves serious refereeing. The central claim of a fast, accessible WQED framework holds up for the core collision model; the two-photon and feedback capabilities are credible but under-validated. I would cite it as a simulation tool, and I would tell the authors to add or point to an external two-photon benchmark in a revision. For a software release of this kind, community use will supply that validation either way.\n\nRecommend: engage with it—send to review.","headline":"A genuinely useful open-source collision-model simulator for waveguide QED; the two-photon and feedback sectors are the least externally validated part, but the core single-photon dynamics are solid.","tokens_in":19659,"tokens_out":4046,"would_cite":true,"duration_ms":35289,"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":"This paper presents WaveguideQED.jl, an open-source Julia framework that simulates waveguide quantum electrodynamics with full two-photon traveling wavefunctions and non-Markovian delayed feedback, using the collision model with…","keywords":["waveguide quantum electrodynamics","collision model","time-bin discretization","non-Markovian feedback","two-photon wavefunctions","numerical simulation","quantum emitter","matrix-free operators"],"falsifier":"Take the two-photon scattering setup and insert a narrow spectral filter between the emitter and the waveguide so that the effective coupling strength varies strongly across the pulse bandwidth; simulate the same problem with an independent spectrally resolved calculation. If the two-photon wavefunctions differ significantly, the flat-coupling assumption that the framework rests on would be shown to be load-bearing.","tokens_in":18814,"feed_emoji":"⚛️","tokens_out":11486,"duration_ms":96118,"temperature":0.7,"pith_summary":"WaveguideQED.jl is an open-source Julia package that turns the collision model of light–matter interaction into a fast simulation tool for waveguide quantum electrodynamics. The traveling field is sliced into discrete time bins that interact one at a time with a localized quantum system, and the full quantum state of up to two photons is kept rather than reduced to expectation values. The paper shows the framework reproduces analytic single-photon scattering, captures the entangled two-photon output from an emitter, handles multiple waveguide channels, and simulates a mirror-induced delayed feedback where a phase of $\\pi$ traps the emitter excitation. The result is that non-Markovian and multi-photon waveguide physics can be studied routinely, with simulations orders of magnitude faster than a naive sparse-matrix implementation and memory use independent of the number of time bins.","feed_headline":"Open-source simulator treats two-photon waveguide QED with feedback","feed_subtitle":"Collision-model code keeps full photon wavefunctions and runs orders of magnitude faster than sparse-matrix methods.","key_machinery":"The central mechanism is the discrete time-bin formalism of the collision model: the waveguide is divided into $N$ bins of width $\\Delta t$, with annihilation and creation operators $w_k$, $w_k^\\dagger$ satisfying $[w_j,w_k^\\dagger]=\\delta_{jk}$, and the interaction Hamiltonian is constant within each bin, so evolution proceeds as a sequence of unitary steps $U_n=\\exp(-\\frac{i}{\\hbar}\\int H_{\\mathrm{int}}dt')$. The paper couples this picture with two implementation techniques: a custom basis that keeps only the vacuum, single-excitation, and two-excitation amplitudes of the waveguide, and matrix-free operator kernels that perform the action of $w_k$ without constructing a matrix, combined through lazy operators that delay all summation, product, and tensor-product algebra until the operator is applied to a state. This makes the time-dependent interaction $i\\hbar\\sqrt{\\gamma/\\Delta t}(\\sigma^\\dagger w_k - \\sigma w_k^\\dagger)$ cheap to evaluate at every step and keeps memory usage constant in the number of bins.","core_discovery":"The paper's central claim is that the quantum collision model, discretized into time-bin modes, can serve as the basis of a general, efficient, and open-source framework for waveguide QED. The key demonstrations are: a custom excitation-restricted basis that represents the two-photon waveguide state with $\\sim N^2/2$ amplitudes instead of an exponentially large product space; waveguide operators implemented as matrix-free kernel functions and combined through lazy operator algebra, so that no sparse matrix is allocated for each time step; a single-photon scattering result that converges to the input–output equations of motion as the time step shrinks; a scattered two-photon state that is genuinely entangled and not a product of single-photon pulses; and a delayed-feedback mirror simulation in which the emitted field returns with a phase and can either hasten decay or trap the excitation at a steady state. These examples support the conclusion that the package offers a complete description of the traveling photon wavefunction in regimes where most existing approaches give only observables or product states.","pith_inferences":["If the framework's flat-coupling assumption is accepted, it becomes a convenient testbed for few-photon entanglement measures and for the design of photon-photon gates, because the complete two-photon wavefunction is available rather than only expectation values.","The same time-bin machinery could be adapted to other local interaction problems with engineered delays, such as electron transport through quantum dots, since the collision picture is not specific to photons.","A direct way to stress-test the method is to apply the convergence protocol of the appendix to a multi-emitter or giant-emitter layout, where delays are comparable to the pulse width; the paper has not yet reported such a benchmark."],"forward_implications":["A researcher can set up and run a single- or two-photon waveguide scattering simulation in minutes, and the results match the analytic input–output equations of motion in the continuum limit.","Delayed feedback from mirrors or distant emitters can be simulated while retaining the full photon wavefunction, making effects such as excitation trapping directly visible.","The waveguide operators combine with arbitrary local systems from the host quantum optics package, so multi-level emitters, cavities, and other local Hamiltonians require no new derivations.","Benchmarks show orders-of-magnitude speedups over preallocating sparse matrices, with memory usage independent of the number of time bins, so fine discretizations are practical.","The same basis and kernel structure can be extended to three- or four-photon states at predictable cost, and to GPU execution."],"supporting_citations":[{"why":"Supplies the collision-model foundation that the time-bin formalism and the framework are based on.","marker":"[40]"},{"why":"Provides the time-bin discretization and the recursive unitary used to derive the single-photon scattering solution that the code reproduces.","marker":"[41]"},{"why":"The host quantum optics package whose basis, operators, and solver infrastructure the framework extends.","marker":"[45]"},{"why":"Lazy operator machinery that the paper uses to combine matrix-free waveguide kernels into general Hamiltonians.","marker":"[49]"},{"why":"Collision model for non-Markovian quantum trajectories that underlies the delayed-feedback mirror Hamiltonian in Section 4.2.","marker":"[50]"},{"why":"Space-discretized waveguide model used as the performance benchmark and comparison for the feedback simulation.","marker":"[14]"}],"fun_headline_variants":["Collision-model simulator handles delayed-feedback waveguide QED","Efficient Julia framework for two-photon non-Markovian QED","Open-source waveguide QED code preserves photon wavefunctions","Fast two-photon scattering with memory effects in Julia"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the emitter couples equally strongly to every photon frequency, which is what lets the field be split into independent time bins that interact one at a time; if a problem has a strongly frequency-dependent coupling, this time-bin collision picture no longer applies.","fun_headline_variants_meta":{"raw":{"variants":["Collision-model simulator handles delayed-feedback waveguide QED","Efficient Julia framework for two-photon non-Markovian QED","Open-source waveguide QED code preserves photon wavefunctions","Fast two-photon scattering with memory effects in Julia"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000266,"raw_usage":{"total_tokens":1585,"prompt_tokens":893,"completion_tokens":692,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":509,"completion_tokens_details":{"reasoning_tokens":623}},"tokens_in":509,"tokens_out":692,"duration_ms":6828,"temperature":1.0,"reasoning_tokens":623,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:13:29.080141+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the two-photon scattering setup and insert a narrow spectral filter between the emitter and the waveguide so that the effective coupling strength varies strongly across the pulse bandwidth; simulate the same problem with an independent spectrally resolved calculation. If the two-photon wavefunctions differ significantly, the flat-coupling assumption that the framework rests on would be shown to be load-bearing.","supporting_citations":[{"cited_title":"Collision models in quan- tum optics","cited_arxiv_id":null,"evidence_quote":"Supplies the collision-model foundation that the time-bin formalism and the framework are based on."},{"cited_title":"Photon-photon interactions in dy- namically coupled cavities","cited_arxiv_id":null,"evidence_quote":"Provides the time-bin discretization and the recursive unitary used to derive the single-photon scattering solution that the code reproduces."},{"cited_title":"QuantumOp- tics.jl: A Julia framework for simulating open Accepted in Quantum 2025-04-10, click title to verify. Published under CC-BY 4.0. 14 quantum systems","cited_arxiv_id":null,"evidence_quote":"The host quantum optics package whose basis, operators, and solver infrastructure the framework extends."},{"cited_title":"LazyOperators in QuantumOptics.jl","cited_arxiv_id":null,"evidence_quote":"Lazy operator machinery that the paper uses to combine matrix-free waveguide kernels into general Hamiltonians."},{"cited_title":"Collision model for non- Markovian quantum trajectories","cited_arxiv_id":null,"evidence_quote":"Collision model for non-Markovian quantum trajectories that underlies the delayed-feedback mirror Hamiltonian in Section 4.2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Space-discretized waveguide model used as the performance benchmark and comparison for the feedback simulation."}],"review_version":1}