{"id":"f14f77c3-5be5-432b-bea4-b06bf63977b8","arxiv_id":"2507.16549","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Direct imaging of excited-state population during subradiant decay in a cold 174Yb cloud reveals decay scaling with optical depth that matches scattered-light data and two-level dipole simulations.","lead":"This paper reports a new way to directly measure the number and location of excited atoms in a cold cloud of ytterbium atoms while it stays dim, or subradiant, and shows the decay time matches both scattered-light measurements and simulations. The method gives a population-based view of subradiance, potentially useful for studying where light gets trapped in dense atomic ensembles.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"At high optical depth, the fitted subradiant decay time exceeds the reliable observation window; the reported slope may be set by the background correction rather than by the decay dynamics.","rationale":"The reader identified the background/density correction as the weakest assumption, and the manuscript itself flags this limitation. My stress-test sharpens the concern into a quantitative issue: the quoted decay times at the largest optical depths are longer than the reliable temporal window, so the linear slope cannot be confirmed from resolved decays alone and may instead be inherited from the baseline model. This does not require accusing the authors of any error; it means the central claim is not yet checkable from the available material. The paper has genuine independent support: the population and intensity observables agree in the time-resolved comparison of Fig. 4, the intensity dependence in Fig. 3(a) is consistent with earlier subradiance results, and the coupled-dipole prediction is a standard model with no free parameters beyond optical depth. Those elements make the claim plausible and explain why outright rejection is not warranted. However, the linear slope is the headline result, and it is precisely the quantity most exposed to the finite-window and baseline issue. The reader's conditional verdict remains appropriate: acceptance should await the fitting details, the supplemental material, and either public data or the synthetic-dataset check described above. I therefore recommend no change to the reader's verdict.","tokens_in":8725,"tokens_out":2725,"duration_ms":34739,"concrete_test":"Regenerate the population-decay datasets for each b0,g in Fig. 2 as synthetic decays with known τ = 0.367 b0,g, add Gaussian noise and a density-dependent background offset matching the stated contrast trend, truncate the data at tΓ = 10, and apply the same fitting procedure with a floating baseline. If the recovered slope differs from 0.367 by more than the reported 0.04 uncertainty, the finite-window and baseline effects alone can explain the measured slope. As a simpler check, re-fit the real data with the baseline fixed to an independently measured late-time value instead of letting it float; if the slope moves outside the stated error bars, the claim depends on the background model.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is the linear scaling of the subradiant decay time with optical depth, τsub/τg = 0.34(4) b0,g, quoted as agreeing with a coupled-dipole prediction of 0.367(9) b0,g. The manuscript itself states that reliable population data are restricted to tΓ < 10 and that the infinite-time background offset increases with density, with contrast decreasing as b0,g grows (Fig. 1(d) and discussion). For the highest optical depths used in the slope fit, b0,g ≈ 50 gives τsub/τg ≈ 17, so the fitted decay constant is nearly twice the entire reliable window. Less than one e-fold of the decay is observed before the background-dominated region is reached. A single-exponential fit over less than one e-fold, with a floating baseline that grows with density, is highly sensitive to the background model; a small density-dependent baseline bias could produce an artificial linear increase of τsub with b0,g. Because the fitting procedure and simulation details are relegated to an unavailable supplemental and the data are not public, the agreement with theory currently rests on this unverified baseline treatment rather than on directly resolved decay dynamics. This is the load-bearing weak point of the central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The letter presents a depletion-imaging technique on the 1S0–1P1 transition of 174Yb to measure, with time and spatial resolution, the excited-state population in the 3P1 state during the decay of a laser-excited cold atomic cloud. It compares the population decay with single-mode-fiber scattered-intensity measurements at 35 degrees and with coupled-dipole numerical simulations. The central quantitative result is a linear scaling of the late-time subradiant decay time with resonant optical depth, reporting τsub/τg = (0.34 ± 0.04) b0,g for population measurements versus a numerical prediction of (0.367 ± 0.009) b0,g. The authors also report independence of the decay rate from driving intensity and detuning, and observe faster decay at the cloud boundaries.","tokens_in":8980,"tokens_out":6878,"duration_ms":74538,"significance":"The work introduces a genuinely new observable—directly imaged excited-state population—into the study of collective subradiance, and it demonstrates a capability (space-resolved decay) that far-field intensity measurements do not provide. If the slope agreement is correct, it is a valuable quantitative confirmation of two-level coupled-dipole behavior in a system with unit Clebsch-Gordan coefficients. I also credit the authors for using an independent numerical prediction rather than fitting the theory to the data. The main risk is that the key fitting and simulation information is not present in the submitted manuscript, and the high-optical-depth data cover less than one e-fold of the decay.","major_comments":[{"comment":"The claim that the late decay time scales linearly with b0,g is not directly resolvable at the optical depths that most constrain the fit. The authors state that for large optical depths the limited contrast does not allow reliable data for tΓ > 10; for b0,g ≈ 50, the fit value τsub/τg ≈ 17 means the fitted decay constant is almost twice the entire reliable window. In this regime the data cover less than one e-fold, and the infinite-time background (dashed lines in Fig. 1(d)) is density-dependent, so the extracted τsub is highly sensitive to the baseline model. A small density-dependent baseline bias would produce an artificial linear increase of τsub with b0,g. Please provide a quantitative check—for example, restrict the slope fit to b0,g values for which τsub is within the reliable window, or show that the baseline subtraction is validated on independent datasets.","section":"Fig. 1(d) and Fig. 2 (population data)"},{"comment":"The fit function, error model, and numerical simulation are presented only by reference to a Supplemental Material that is not included in the submitted manuscript. The central comparison (0.34(4) versus 0.367(9)) cannot be checked as submitted. Please include, in the manuscript or in the supplement provided to reviewers, the explicit decay model (including the treatment of Σ′_{e,t→∞} as a free parameter per dataset), the uncertainty propagation from the decay fits to the slope, and sufficient coupled-dipole simulation parameters (number of atoms, geometry, cloud profile, and fitting range for τsub) to reproduce the numerical scaling.","section":"Supplemental material references (fitting procedure and simulation)"},{"comment":"The stated slope uncertainty of ±0.04 appears to reflect only the statistical error of the linear fit. The systematic sensitivity of the slope to the choice of fitting window, background model, and the b0,g range included in the fit is not discussed. Please report these systematic checks; without them the claimed 1σ agreement with 0.367(9) is not a meaningful statement of accuracy.","section":"Fig. 2 and slope uncertainty"}],"minor_comments":[{"comment":"The abstract describes the internal level structure as V-type, while the main text calls it Λ-type; please reconcile the terminology.","section":"Abstract and main text"},{"comment":"References 5 and 26 are the same paper (Guerin, Araújo, and Kaiser, PRL 116, 083601) listed twice with different formatting; consolidate them.","section":"References"},{"comment":"The inset of Fig. 4 has a garbled axis label (\"tΓg\") and a typographic apostrophe in Σ‘e; please fix the typography and define all axes explicitly.","section":"Fig. 4"},{"comment":"The ratio image is first defined as Id/Ir, then R(x,y) is introduced as Id/Ir − 1; the same symbol should not be used for both, and the sentence containing \"absorption picture\" should read \"absorption imaging\".","section":"Eq. (2) and surrounding text"},{"comment":"The linear fit to the contrast C as a function of b0 is shown without fit parameters or error bars; report the extracted slope and intercept.","section":"Fig. 1(d) inset"},{"comment":"The scattered-intensity data (black triangles) are not fitted or quantitatively compared; either include a fit and its uncertainty or state explicitly that these points are shown only for qualitative comparison.","section":"Fig. 2"}],"recommendation":"major_revision","confidential_remarks":"I want to emphasize to the editor that the absence of the supplemental material is a blocking issue for this specific claim. The paper's own limitation paragraph is honest, but the central slope may well be controlled by the density-dependent background correction. The authors should be asked to supply the fitting details and a robustness analysis. If those are satisfactory, this could be a strong letter; if not, the central quantitative claim is not supported by the visible data."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a real experimental step forward. They directly image the excited-state population of an optically dense Yb cloud during subradiant decay, something only inferred from photon counting before, and they get spatially-resolved decay maps. The main quantitative claim is that tau_sub scales linearly with optical depth with slope 0.34(4), matching an independent two-level coupled-dipole prediction of 0.367(9). That agreement is real on its face, and the model is not fit to the data. Good.\n\nThe technique itself is borrowed from their prior PRL (depletion imaging on Yb), so the novelty is the application to subradiance and the combined population-plus-intensity comparison. The intensity-independence and detuning-independence checks are solid and add confidence. Citation pattern is fine; they reference their own prior calibration and the relevant subradiance literature. The V/\\Lambda transition inconsistency between the abstract and the full text is a minor typo but should be fixed.\n\nThe soft spot is exactly what the paper flags: the reliable data window is t\\Gamma < 10, and for the highest optical depths used in the slope fit (b0 ~ 50), the fitted tau_sub ~ 17 tau_g, which is nearly twice the entire observation window. That means less than one e-fold of the decay is directly observed before the density-dependent background dominates. They subtract an infinite-time offset that grows with density, and the contrast decreases with b0. If that baseline has a small density-dependent bias, it could create a spurious linear slope. The fitting procedure, error model, and simulation details are all in a supplemental that is not provided, and the data are not public. That makes the central agreement hard to verify. The reader's conditional verdict is fair.\n\nThat said, the paper is openly honest about the limitation, and the scaling law itself is not new—it has been predicted and seen via intensity measurements. The new contribution is the observable, and that stands. The slope measurement is suggestive but not definitive until the baseline treatment is shown to be stable across the full range.\n\nRecommendation: send to peer review. A serious referee should ask for the supplemental and a baseline sensitivity analysis. This deserves consideration as a letter after revision; the technique and the spatially-resolved maps are worth publishing even if the scaling claim gets softened.","headline":"New population-based observable for subradiance, but the quantitative scaling claim rests on a background correction that isn't fully demonstrated.","tokens_in":9502,"tokens_out":2004,"would_cite":false,"duration_ms":20742,"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":"Direct imaging of excited-state population in cold 174Yb clouds shows that the subradiant decay time grows linearly with optical depth, with a measured slope of 0.34(4) that matches the two-level coupled-dipole prediction of 0.367(9).","keywords":["subradiance","cold atomic ensembles","depletion imaging","excited-state population measurement","optical depth scaling","ytterbium","collective scattering"],"falsifier":"Re-measure the population decay at a fixed optical depth while reducing the imaging probe intensity enough to suppress radiation pressure and light-assisted collisions. If the extracted late-time decay constant changes by more than the reported uncertainty, the 0.34(4) b0,g slope is an artifact of the background correction rather than intrinsic subradiant dynamics.","tokens_in":8558,"feed_emoji":"⚛️","tokens_out":7530,"duration_ms":80712,"temperature":0.7,"pith_summary":"Subradiant atomic ensembles release stored light much more slowly than single atoms, but experiments have usually watched only the emitted photons, leaving the internal atomic state hidden. This paper reports a depletion-imaging method for cold 174Yb atoms that directly records the excited-state population while the subradiant modes decay, and it compares that population with the simultaneously measured scattered light. The central quantitative result is that the late subradiant decay time scales linearly with resonant optical depth, τsub/τg = 0.34(4) b0,g, in agreement with the two-level coupled-dipole prediction of 0.367(9) b0,g and with the intensity-based measurement. By seeing the same timescale in an observable that carries coherence information and one that does not, the paper establishes a direct, spatially resolved signature of subradiance in the atomic state itself.","feed_headline":"Subradiant decay time grows with optical depth, imaged directly","feed_subtitle":"Population and scattered-light measurements agree, confirming a clean two-level description of subradiant decay.","key_machinery":"The central object is the depletion image ratio $R(x,y)=I_d(x,y)/I_r(x,y)-1$ between a cloud with excited-state population and a reference cloud, which maps the excited-state column density. The identity $\\Sigma_e = \\frac{1}{\\sigma_{\\mathrm{sc,b}}}\\iint R(x,y)\\,dx\\,dy$ converts that two-dimensional image into the total number of atoms in the $^3P_1$ state. The time resolution comes from a 200 ns probe pulse on the broad 399 nm transition, far shorter than the 1 μs natural lifetime of the narrow intercombination transition, so the excited population is effectively frozen during each snapshot. Repeated at a chosen delay after the excitation pulse, these images give the full decay curve $\\Sigma_e(t)$.","core_discovery":"The paper claims that the subradiant decay time of an optically dense cold atomic cloud can be measured directly from the time-resolved excited-state population, and that this population observable is quantitatively described by a simple two-level coupled-dipole model in the weak-driving limit. Using the broad and narrow transition pair of 174Yb, the authors image the population of atoms in the 3P1 state after a strong resonant excitation pulse; integrating the image ratio gives Σe(t). At late times, they extract τsub and find τsub/τg = 0.34(4) b0,g from the population fits, consistent with scattered-intensity measurements and with the numerical prediction τsub/τg = 0.367(9) b0,g. The same technique shows that cloud boundaries decay faster than the center, and that subradiant population (but not the decay time) grows with driving intensity, corroborating earlier saturation-pumping results.","pith_inferences":["If the early-time difference between intensity and population reflects coherent versus incoherent content, that difference could be used as a time-resolved measure of the transition from superradiant to subradiant dynamics.","Comparing τsub measured in different spatial regions with the local optical depth b(r) would test whether subradiant decay is governed by a local-density picture or by global modes of the whole cloud.","The technique should transfer to strontium and other alkaline-earth-like atoms, and to any species with a narrow cooling transition plus a broad probe transition, enabling population-resolved collective-dynamics studies where alkali hyperfine structure spoils the two-level comparison."],"forward_implications":["Population and scattered-intensity measurements give the same subradiant decay time at late times, showing both are valid probes and that late-time subradiant emission is effectively incoherent.","The linear scaling of τsub with optical depth, reproduced by the two-level coupled-dipole model, confirms that the 1S0–3P1 transition in 174Yb behaves as a clean quantum version of a classical dipole.","Spatially resolved population imaging shows slower decay in the high-optical-depth center than at the edges, opening the way to direct studies of the spatial structure of collective modes.","The decay time is independent of driving intensity and detuning, while the subradiant population grows with intensity, confirming the pumping mechanism that populates long-lived states without changing their lifetime.","The method can track light transport and localization inside dense ensembles by following the excited-state population in situ."],"supporting_citations":[{"why":"Demonstrates the depletion imaging technique on which the population measurements are based.","marker":"[21]"},{"why":"Supplies the coupled-dipole numerical prediction for the subradiant scaling and the prior intensity-based observation the paper reproduces.","marker":"[5]"},{"why":"Earlier cold-atom subradiance experiment using scattered-photon detection, providing the protocol and comparison for the intensity channel.","marker":"[6]"},{"why":"Establishes that pumping at saturation enhances the population of subradiant modes, motivating the chosen excitation intensity.","marker":"[27]"},{"why":"Gives the relation between scattered intensity and two-atom coherences used to contrast the information content of intensity versus population.","marker":"[22]"},{"why":"Motivates measuring population rather than only optical coherence by showing that entanglement witnesses require population fluctuations.","marker":"[23]"}],"fun_headline_variants":["Imaging cold atom population reveals subradiant decay rates","Subradiant lifetimes extracted from direct population imaging","Cold atom cloud: subradiant decay mapped by imaging","Optical depth sets subradiant decay, direct imaging shows","Population snapshots capture subradiant dynamics in Yb"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The population measurement assumes that density differences between the reference and depleted clouds caused by the imaging probe (radiation pressure and light-assisted collisions) are either negligible or completely removed by subtracting the infinite-time background, which the authors state limits reliable data to tΓ < 10.","fun_headline_variants_meta":{"raw":{"variants":["Imaging cold atom population reveals subradiant decay rates","Subradiant lifetimes extracted from direct population imaging","Cold atom cloud: subradiant decay mapped by imaging","Optical depth sets subradiant decay, direct imaging shows","Population snapshots capture subradiant dynamics in Yb"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1419,"prompt_tokens":812,"completion_tokens":607,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":428,"completion_tokens_details":{"reasoning_tokens":541}},"tokens_in":428,"tokens_out":607,"duration_ms":7120,"temperature":1.0,"reasoning_tokens":541,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:06:52.609043+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure the population decay at a fixed optical depth while reducing the imaging probe intensity enough to suppress radiation pressure and light-assisted collisions. If the extracted late-time decay constant changes by more than the reported uncertainty, the 0.34(4) b0,g slope is an artifact of the background correction rather than intrinsic subradiant dynamics.","supporting_citations":[{"cited_title":"Glicenstein, A","cited_arxiv_id":null,"evidence_quote":"Demonstrates the depletion imaging technique on which the population measurements are based."},{"cited_title":"Guerin, M","cited_arxiv_id":null,"evidence_quote":"Supplies the coupled-dipole numerical prediction for the subradiant scaling and the prior intensity-based observation the paper reproduces."},{"cited_title":"Ferioli, A","cited_arxiv_id":null,"evidence_quote":"Earlier cold-atom subradiance experiment using scattered-photon detection, providing the protocol and comparison for the intensity channel."},{"cited_title":"Cipris, N","cited_arxiv_id":null,"evidence_quote":"Establishes that pumping at saturation enhances the population of subradiant modes, motivating the chosen excitation intensity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the relation between scattered intensity and two-atom coherences used to contrast the information content of intensity versus population."},{"cited_title":"Rosario, A","cited_arxiv_id":null,"evidence_quote":"Motivates measuring population rather than only optical coherence by showing that entanglement witnesses require population fluctuations."}],"review_version":1}