{"id":"6900b7dc-8fdd-44d3-bf03-5e7b1774b74f","arxiv_id":"2501.09261","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A photonic-lattice experiment observes non-Markovian, oscillatory decay of a simulated quantum emitter, with flat-band Lieb lattices showing the effect even at weak coupling.","lead":"Researchers used laser-written photonic lattices to simulate a quantum emitter decaying into structured reservoirs, observing oscillations in the emitter's excitation that signal non-Markovian behavior. The flat band of a Lieb lattice produced the strongest non-Markovian signatures even at weak coupling, offering a platform to study light-matter interactions in engineered materials.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The λ→z mapping is calibrated on bulk participation ratios, not on the QE-site amplitude, and V2(λ) is uncharacterized; the Fig. 4(d) revival could be a coupling sweep rather than non-Markovian re-excitation.","rationale":"I read the paper in good faith: the experimental platform is plausible, the square-lattice control behaves as expected, and the bulk Lieb B/C participation-ratio data do support a λ↔z correspondence at the level of R. The most load-bearing condition for the headline claim is that the λ-scan realizes a fixed-Hamiltonian time evolution for the local QE amplitude. That condition is not secured because the calibration observable R is spatially integrated and phase-insensitive, and because V2(λ), the coupling whose constancy is explicitly invoked in the 'equal V2' comparison, is not characterized over the scan range. I agree with the reader's identified weak spot in substance, but not with the claim that no Lieb validation exists: Fig. 3(e) calibrates the mapping for Lieb bulk B and C excitations. The missing piece is the QE-site observable and the wavelength dependence of V2. A fixed-wavelength length series would settle whether the 800 nm revival is genuinely temporal. This does not move the reader's CONDITIONAL verdict; it sharpens the condition under which acceptance should occur.","tokens_in":12338,"tokens_out":10458,"duration_ms":110974,"concrete_test":"Fabricate two additional Lieb samples of different lengths (e.g., L=3 cm and L=7 cm) with the same writing parameters and measure the QE-waveguide power fraction as a function of propagation distance at fixed λ=730 nm; then compare this z-series with the λ-scan in Fig. 4(d) under the claimed mapping λ=λ0+αz. If the two curves do not collapse onto the same cQE(z), the wavelength-to-time mapping is not valid for the QE observable and the observed revival at λ≈800 nm cannot be attributed to non-Markovian dynamics. As a supporting check, report V2(λ) from the same coupler calibration used for Vx(λ) over 600–800 nm.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that a Lieb flat band produces striking non-Markovian dynamics at equal V2—requires the wavelength scan to be equivalent to propagation at fixed Hamiltonian. The calibration in Fig. 3(e) fits λ=λ0+αz to the participation ratio R for square and Lieb bulk B/C excitations; that is real but incomplete support. R is a phase-insensitive, spatially integrated observable, while the claimed non-Markovian signature is the coherent revival of the QE waveguide in Fig. 4(d). The text reports Vx(λ) varying strongly over the scan range (Fig. 3(d)), yet does not report V2(λ) for the QE–lattice gap or the ratio V2/V1 over λ∈{600,800} nm. If V2/V1 drifts with wavelength, the λ sweep changes the coupling regime along the purported time axis: the λ≈800 nm re-population could be a wavelength-tuned resonance with the flat-band compact state rather than a fixed-coupling non-Markovian re-excitation. Thus the Lieb-versus-square comparison at 'equal V2' is not yet secured.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an all-optical waveguide-array experiment intended to emulate the single-excitation decay dynamics of a quantum emitter coupled to two-dimensional structured reservoirs. A femtosecond-written array of evanescently coupled waveguides acts as the reservoir, and one additional waveguide plays the role of the emitter. The authors use a wavelength-scan method in which varying the excitation wavelength λ over 600–800 nm is interpreted as varying an effective propagation distance z (and hence time) through a linear mapping. They compare square and Lieb lattices and report that the Lieb flat band produces stronger non-Markovian signatures, namely re-population of the emitter waveguide and Rabi-like oscillations, even in the weak-coupling regime. The paper also claims to validate the wavelength-scan technique for 2D lattices by comparing measured participation ratios with numerical simulations.","tokens_in":12615,"tokens_out":5287,"duration_ms":56012,"significance":"If the λ-to-z mapping is fully validated, the result would be a valuable experimental platform for simulating structured-reservoir quantum optics, extending photonic-lattice analogue simulations to 2D flat-band reservoirs. The qualitative comparison of square and Lieb lattices, the use of the participation ratio to characterize wave-packet spreading, and the explicit connection to flat-band compact localized states are strengths. The paper is also honest about the single-particle limitation of the platform. The significance is currently compromised by the lack of characterization of the wavelength dependence of the emitter–reservoir coupling and by the fact that the mapping is calibrated on a phase-insensitive bulk observable rather than on the emitter-site observable that carries the central non-Markovian claim.","major_comments":[{"comment":"The linear transformation λ = λ0 + αz is introduced as an 'adjustment' to make discrete simulations match experiment, but α and λ0 are not reported in the main text, and the calibration observable is the participation ratio R for bulk square and Lieb excitations. Fig. 4(d), which contains the central non-Markovian revival claim, is a different observable (fraction of power in the QE waveguide) and a different geometry (Lieb, with A- and B-site coupling). The mapping needs to be validated directly for the QE-site amplitude and for the Lieb geometry, or the quantitative interpretation of the λ axis as a time axis is not secured.","section":"Sec. 3, Fig. 3(e)"},{"comment":"Vx(λ) varies substantially over the scan range, yet V2(λ), the emitter–reservoir coupling, and the ratio V2/V1 are not reported as functions of λ. If V2/V1 drifts with wavelength, the scan sweeps the coupling regime rather than evolving the same Hamiltonian in time; the re-population near λ ≈ 800 nm could then be a wavelength-tuned resonance with the flat-band compact state rather than a fixed-coupling non-Markovian re-excitation. The authors should characterize V2(λ), show that V2/V1 is constant (or apply a rescaling that accounts for drift), and provide a direct simulation of the curves in Fig. 4(d) using the same fitted mapping.","section":"Fig. 3(d) and Fig. 4(d)"},{"comment":"The stated analogy between λ and z requires that a change in wavelength rescale the entire Hamiltonian by a common factor, not alter the normalized couplings. The text only establishes Vx(λ); it does not show that the ratios Vx/V1, Vy/V1, and V2/V1 are wavelength independent. A quantitative statement of the Hamiltonian rescaling, or a demonstration that the non-Markovian signatures persist in constant-coupling propagation experiments at several fixed wavelengths, is needed.","section":"Eq. (1) and following paragraph"},{"comment":"For the Lieb weak-coupling A-site curve, the text asserts that the compact state 'excites back the QE' and that this is observed near λ ≈ 800 nm; however, no numerical simulation of the QE-power curve is shown for this configuration. Adding the simulated QE power versus z/λ using the calibrated α would make the claim falsifiable and would allow the reader to assess whether the position and amplitude of the revival are quantitatively reproduced.","section":"Sec. 4, Fig. 4(d)"}],"minor_comments":[{"comment":"The text should spell 'waveguide array' consistently rather than 'PW A'; 'Lets us consider' should be 'Let us consider'; and 'monotonous' should be 'monotonic'.","section":"Throughout"},{"comment":"The same reference is cited twice as [35] and [37] (Lederer et al.); please consolidate the duplicated entry.","section":"References"},{"comment":"The number of lattice sites N in Eq. (2) is not defined in the main text, and the factor 2√N is not derived; please clarify the normalization and the conditions under which the cosine solution applies.","section":"Eq. (2)"},{"comment":"The output intensity profiles in Fig. 4 would be easier to interpret with color bars or scale bars, and the caption should state which V2/V1 value corresponds to each curve at the central wavelength.","section":"Fig. 4"},{"comment":"The claim that the wavelength-scan technique is validated for 2D lattices would be easier to evaluate if λ0, α, and their uncertainties were given in the main text rather than only in the Supplemental Material.","section":"Sec. 3, calibration text"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the stress-test concern raised by the reader is, in my assessment, the central unresolved issue. The manuscript is generally careful and the experimental dataset is valuable, but the load-bearing temporal interpretation depends on a fitted mapping that is not yet validated for the emitter-site observable in the Lieb geometry, and the wavelength dependence of V2 is not characterized. These points are fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a waveguide-array experiment claiming to observe non-Markovian decay of a simulated quantum emitter coupled to 2D structured reservoirs, with a Lieb flat band enhancing the non-Markovianity. The result is plausible and qualitatively convincing; the flat-band localization and the revival of the emitter are clearly visible in the data. But the central quantitative claim—that the comparison between square and Lieb is made at equal emitter-reservoir coupling V2—is not properly supported, because the paper never characterizes V2 as a function of wavelength.\n\nWhat's new: the experimental implementation itself. The theory for non-Markovian dynamics in 2D lattices is prior work (González-Tudela and Cirac), and flat-band enhanced coupling has been predicted. The new piece is the observation of these effects in fabricated square and Lieb lattices, using a wavelength scan as an effective propagation distance. The bulk characterization of the wavelength-to-distance mapping (Fig. 3e) is solid work—they show the participation ratio for square and Lieb bulk excitations and get a good fit to a linear λ = λ0 + αz after matching to numerics. They also characterize Vx(λ) with a quadratic fit. That is real, reproducible effort.\n\nThe soft spot, as the stress-test note says, is that this mapping is calibrated on a phase-insensitive bulk observable (participation ratio) and does not involve the emitter. The QE-waveguide coupling V2 is not reported versus wavelength. Since Vx varies substantially across the 600–800 nm scan, V2 almost certainly does too. So a wavelength sweep changes the coupling regime along the purported time axis. The revival near 800 nm in Fig. 4(d) could be a wavelength-tuned resonance with the flat-band compact state rather than a fixed-coupling non-Markovian re-excitation. The authors should measure a coupler at the relevant QE-lattice gap and report V2/V1 versus λ. Without that, the “equal V2” comparison between square and Lieb is not secured. Also, the paper shows no error bars on the emitter power curves, and the fitting parameters α and λ0 are relegated to the SM.\n\nThat said, this is not a fatal flaw. The qualitative story—square weak coupling decays, strong coupling oscillates, Lieb weak coupling revives—is robust and is what makes the paper interesting. The mapping criticism is a significant but fixable gap. If the authors characterize V2(λ) and either validate the mapping for the emitter observable or at least discuss the coupling sweep explicitly, the paper becomes a solid demonstration. Who benefits: people working on photonic simulators of open quantum systems, flat-band physics, and non-Markovian quantum optics. It deserves a serious referee; I would send it out, with the expectation that the revision addresses the V2(λ) issue directly. I would cite it for the platform demonstration even with the caveat.","headline":"Plausible and visually convincing observation of flat-band-enhanced non-Markovian emitter dynamics in a photonic waveguide array, but the central 'equal V2' claim needs direct characterization of V2(λ) before publication.","tokens_in":13127,"tokens_out":4397,"would_cite":true,"duration_ms":47274,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The flat band of a Lieb lattice drives an artificial emitter into non-Markovian decay, with oscillatory re-excitation, as seen in photonic waveguide arrays.","keywords":["non-Markovian dynamics","photonic waveguide arrays","Lieb lattice","flat band","structured photonic reservoir","quantum emitter analog","wavelength scan method","light-matter coupling"],"falsifier":"Fabricate two Lieb samples with different physical lengths, or cleave one sample, and image the output at fixed wavelength for several propagation distances; if the re-excitation pattern in $z$ does not match the wavelength-scan curve under the same $\\lambda_0$ and $\\alpha$, the effective-time interpretation fails. Also measure $V_2(\\lambda)$ across the scan range; a strong wavelength dependence would let coupling changes mimic genuine temporal oscillations.","tokens_in":12161,"feed_emoji":"⚛️","tokens_out":7538,"duration_ms":73223,"temperature":0.7,"pith_summary":"The paper reports a tabletop optical analog of a quantum emitter decaying into a structured reservoir, and claims that the flat band of a Lieb lattice turns an otherwise Markovian weak-coupling decay into non-Markovian dynamics with oscillatory re-excitation. In the analog, a single waveguide plays the emitter, the surrounding lattice plays the reservoir, and light propagation along the waveguide plays time. The authors compare square and Lieb reservoirs, find quasi-exponential decay for the square lattice in weak coupling and Rabi-like oscillations in strong coupling, and find that the Lieb lattice shows clear memory effects even when the emitter-reservoir coupling is weak. They trace the effect to the flat band's vanishing group velocity and large density of states, which hold excitation near the emitter and feed it back. If the claim stands, flat-band reservoirs become a practical way to get strong-coupling-style behavior without requiring large coupling strengths.","feed_headline":"Flat-band lattice makes an emitter radiate with memory","feed_subtitle":"Waveguide experiments show oscillatory re-excitation of a quantum-emitter analog coupled to a Lieb lattice.","key_machinery":"The key object is the photonic waveguide array as a quantum-optical analog: one evanescently coupled waveguide acts as the emitter, the surrounding lattice acts as the reservoir, and the discrete linear Schr\\\"odinger equation maps propagation distance $z$ to the emitter's evolution time. The experimental enabling device is the wavelength scan: because inter-waveguide coupling increases with wavelength, sweeping $\\lambda$ is treated as a linear sweep of effective distance through $\\lambda = \\lambda_0 + \\alpha z$, so a single 5-cm sample yields a continuous time trace. On the reservoir side, the Lieb lattice provides a flat band at zero energy, whose zero group velocity and concentrated density of states keep the emitted excitation localized near the emitter and send part of it back, producing the oscillatory re-excitation that the paper identifies as non-Markovian.","core_discovery":"The central claim is that the spectral structure of a two-dimensional reservoir, not the bare coupling strength, controls whether an emitter's decay is Markovian, and that a flat band is a particularly strong controller. For a Lieb lattice, whose bands include a dispersionless band at zero energy, the paper argues and shows that a weakly coupled emitter repeatedly exchanges excitation with a compact localized state residing on the A and C sublattices, producing oscillations in the emitter's remaining power rather than a smooth decay. These oscillations are the observed signature of non-Markovianity: the emitter's current state depends on its past because the reservoir stores and returns the excitation. The experiment demonstrates this in an all-optical waveguide-array setup, where varying the excitation wavelength from 650 to 800 nm acts as a proxy for increasing propagation distance, and hence for advancing time. On the author's terms, this is an observation of non-Markovian radiative phenomena in structured photonic lattices, with the Lieb lattice's flat band the enhancing ingredient.","pith_inferences":["A direct test that would separate the flat-band mechanism from calibration artifacts is to measure the emitter-reservoir coupling $V_2(\\lambda)$ across the scan range; if it varies strongly with wavelength, part of the observed oscillation could come from changing coupling rather than from genuine temporal memory.","The wavelength-to-time mapping was validated on a square lattice and then assumed for the Lieb geometry; re-checking the mapping with a fixed wavelength on several physical lengths would close that gap and is within reach of the same fabrication technique.","The single-particle linear platform cannot test whether these memory signatures survive photon-photon interactions; observing non-Markovian effects with interacting photons would require a nonlinear or strongly correlated extension not present in this experiment."],"forward_implications":["Weak coupling to a flat-band reservoir can mimic strong-coupling dynamics, so coherent exchange between an emitter and a reservoir does not require a large emitter-reservoir coupling constant.","The wavelength-scan method turns one fabricated sample into a full set of effective propagation times, making 2D reservoir studies feasible without many length-controlled samples.","Because the analog is linear and single-particle, it can be extended to many indistinguishable emitters coupled to a common structured reservoir, opening the same non-Markovian physics to collective effects.","Lieb-lattice compact localized states are the memory-storing modes in this experiment; reservoirs engineered around other localized or dispersionless bands should show similar re-excitation dynamics."],"supporting_citations":[{"why":"Predicted non-Markovian decay of an emitter coupled to a 2D square-lattice reservoir; the experiment's square-lattice weak and strong coupling behavior is checked against this.","marker":"[11]"},{"why":"Introduced the flat band and compact localized states of the Lieb lattice, the reservoir whose memory effect is central here.","marker":"[17]"},{"why":"Independent observation of Lieb-lattice flat-band states; supports the flat-band phenomenology used for the A and C site excitations.","marker":"[18]"},{"why":"Showed that waveguide arrays can simulate an emitter coupled to a reservoir, the methodological foundation of this experiment.","marker":"[31]"},{"why":"Defined the weak and strong coupling regimes and Rabi-oscillation signatures used to classify Markovian versus non-Markovian dynamics.","marker":"[33]"},{"why":"Supplemental Material supplies the coupling constants, the $\\lambda = \\lambda_0 + \\alpha z$ calibration, and numerical details on which the interpretation of the wavelength scan rests.","marker":"[36]"},{"why":"Femtosecond laser-writing technique used to fabricate the square and Lieb waveguide arrays.","marker":"[38]"},{"why":"Showed that Lieb-lattice excitation is input-site dependent, which determines the choice of A and B sites for the emitter coupling.","marker":"[39]"}],"fun_headline_variants":["Flat-band lattice imprints memory on emitter decay","Photonic lattice reveals non-Markovian radiation","Lieb lattice flat band drives emitter memory","Structured lattice shows emitter re-excitation","Non-Markovian decay observed via flat band"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that scanning the excitation wavelength is exactly equivalent to letting the same system evolve for a longer time, and that the linear calibration established for the square lattice remains valid for the Lieb lattice.","fun_headline_variants_meta":{"raw":{"variants":["Flat-band lattice imprints memory on emitter decay","Photonic lattice reveals non-Markovian radiation","Lieb lattice flat band drives emitter memory","Structured lattice shows emitter re-excitation","Non-Markovian decay observed via flat band"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000705,"raw_usage":{"total_tokens":3142,"prompt_tokens":869,"completion_tokens":2273,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":485,"completion_tokens_details":{"reasoning_tokens":2202}},"tokens_in":485,"tokens_out":2273,"duration_ms":16329,"temperature":1.0,"reasoning_tokens":2202,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:07:09.102365+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Fabricate two Lieb samples with different physical lengths, or cleave one sample, and image the output at fixed wavelength for several propagation distances; if the re-excitation pattern in $z$ does not match the wavelength-scan curve under the same $\\lambda_0$ and $\\alpha$, the effective-time interpretation fails. Also measure $V_2(\\lambda)$ across the scan range; a strong wavelength dependence would let coupling changes mimic genuine temporal oscillations.","supporting_citations":[{"cited_title":"Gonz´ alez-Tudela and J","cited_arxiv_id":null,"evidence_quote":"Predicted non-Markovian decay of an emitter coupled to a 2D square-lattice reservoir; the experiment's square-lattice weak and strong coupling behavior is checked against this."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced the flat band and compact localized states of the Lieb lattice, the reservoir whose memory effect is central here."},{"cited_title":"Crespi, F","cited_arxiv_id":null,"evidence_quote":"Showed that waveguide arrays can simulate an emitter coupled to a reservoir, the methodological foundation of this experiment."},{"cited_title":"Calaj´ o, F","cited_arxiv_id":null,"evidence_quote":"Defined the weak and strong coupling regimes and Rabi-oscillation signatures used to classify Markovian versus non-Markovian dynamics."},{"cited_title":"Szameit, D","cited_arxiv_id":null,"evidence_quote":"Femtosecond laser-writing technique used to fabricate the square and Lieb waveguide arrays."}],"review_version":1}