{"id":"2fa85751-11a3-4a4c-8693-262c61df848d","arxiv_id":"2607.29515","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A single atom's resonance fluorescence interferes with an external coherent probe beam, and the interference visibility gives a non-destructive, time-resolved temperature measurement.","lead":"A single rubidium atom trapped in a laser can brighten or dim a weak resonant laser beam, depending on the phase of its own fluorescence. The size of this effect gives a fast, non-destructive readout of the atom's temperature, demonstrated with 4% precision at about 30 microkelvin.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Thermometry estimator may fold spatial amplitude correlations and anharmonicity into temperature bias beyond the reported 4%","rationale":"The paper's central demonstration—single-atom resonance fluorescence interfering with a coherent beam—is credible and supported by the observed phase-dependent count-rate modulation. The thermometry claim, however, rests on a specific model of how the interference visibility depends on atomic temperature. The weakest point is not a typo or a missing derivation; it is that the ensemble average is factorized into an amplitude factor and a pure phase-dephasing factor. In an optical dipole trap, the atom's scattering amplitude into the collection mode is not constant over the region explored at the reported temperatures, and the trap is not perfectly harmonic. Both effects change the visibility without changing the Gaussian phase variance assumed in Eq. 8. This means the estimator can be biased even if the fit to the histograms is excellent, and the bias is not captured by the R&R endpoint comparison because that comparison uses an interpolated heating model rather than independent intermediate-temperature measurements. The reader's weakest_assumption identified the harmonic-Boltzmann Gaussian-phase assumption; the concern here is broader and more concrete: even within a thermal distribution, the amplitude–position correlation invalidates the factorization. This reinforces the CONDITIONAL verdict: the experimental observation is sound, but the thermometric calibration needs an explicit test or model extension before the 4% accuracy claim can be taken at face value.","tokens_in":10082,"tokens_out":11589,"duration_ms":138553,"concrete_test":"Compute the exact thermal average A(T) = |∫ d^3r P_T(r) sqrt(eta(r)|Omega_pump(r)|^2) exp(i 2π(dY-dZ)/lambda)| using the independently characterized FORT potential (852 nm, 12 mW, 1.32 um waist) and collection-mode profile, for temperatures 18–133 uK and with P_T obtained from the actual anharmonic potential. Compare A(T)/(sqrt(overline{Nsc}(T) overline{Npr})) with exp[-k_B T (2π/lambda)^2/(m omega^2)] used in Eq. 8. If the ratio deviates by more than 4% over the studied range, the calibration bias exceeds the claimed precision, and the thermometry claim would need re-analysis with the full spatial average.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equations 2–4 and estimator (8) replace the ensemble average of sqrt(Nsc(r) Npr) exp(i phi_cm(r)) by sqrt(overline{Nsc} overline{Npr}) exp(-sigma_phi^2/2), assuming that only the phase phi_cm is randomized while the scattering amplitude is position-independent. In the real FORT, however, Nsc(r) depends on position through collection efficiency and light shifts, and at 30–133 uK the rms displacement is a substantial fraction of the 1.32 um waist. Thus amplitude–position correlations and trap anharmonicity both reduce the interference visibility independently of temperature. Equation 8 then attributes all visibility loss to a Gaussian phase variance, biasing the inferred T. The R&R comparison anchors only the 18 uK and 133 uK endpoints and interpolates between them with a scattering model, so it cannot detect a mid-range systematic bias. The quoted 4% uncertainty is the fit covariance only, not this calibration error.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports an experiment in which a single 87Rb atom in a far-off-resonance optical dipole trap is simultaneously illuminated by a resonant pump beam and a weak resonant probe beam. The collected probe-mode light shows first-order interference that depends on the relative phase between the probe and the atom's resonance fluorescence, confirming the predictions of Goncalves et al. The authors develop a statistical model of the photon-count histograms, including an arcsine distribution convolved with technical noise, and use the fitted visibility to infer the atomic center-of-mass temperature. They report temperature uncertainties of about 4% for ~30 μK temperatures with ~200 μs time resolution, using 1200 atoms with 80 1-ms exposures each, and compare the resulting temperatures with release-and-recapture thermometry at the endpoints.","tokens_in":10370,"tokens_out":7178,"duration_ms":84909,"significance":"If the method works as claimed, it provides a non-destructive, time-resolved thermometer for single trapped atoms and ions, with better time resolution than destructive methods and without losing the atom. This would be useful for optimizing cooling, feedback control, and quantum information experiments. The central interference observation is interesting and the statistical model is clearly presented. The paper also gives a concrete, falsifiable prediction connecting interference visibility to temperature, using independently measured scattering rates, probe rates, and trap frequency rather than a fitted temperature constant. However, the manuscript contains several dimensional inconsistencies in the printed formulas and does not quantify a potentially important systematic bias from position-dependent scattering amplitudes and trap anharmonicity; these issues need to be addressed before the accuracy claim is fully supported.","major_comments":[{"comment":"The steady-state density-matrix elements are printed as rho_ee = |Omega|^2/Gamma_0 and rho_eg = i Omega/Gamma_0. As written, rho_ee has units of rate (s^-1) rather than dimensionless population, and rho_eg is likewise dimensionally inconsistent. Similarly, the phase variance is stated as sigma_phi^2 = 2 k_B T/(m omega^2), which has units of length^2; since phi_cm = 2pi(dY-dZ)/lambda, the variance must contain (2pi/lambda)^2. This is not a purely cosmetic issue: Eq. (8) includes lambda^2/(4pi^2), which is inconsistent with the printed sigma_phi^2. Please correct these formulas and re-derive Eqs. (2) and (8) so that the numerical factors can be checked.","section":"Eqs. (1)-(2) and text after Eq. (2)"},{"comment":"The visibility model averages only the phase phi_cm and takes the amplitude prefactor sqrt(Nsc Npr) outside the average. However, the paper itself notes in the discussion of Fig. 3 that the mean count shifts as the atom heats, attributed to reduced light shifts and collection efficiency at larger displacements. At the temperatures considered, the rms displacement is a substantial fraction of the 1.32 um waist, so position-dependent scattering amplitude and trap anharmonicity can reduce the visibility independently of temperature. Equation (8) attributes all visibility loss to the Gaussian phase variance, which would bias the inferred temperature. The quoted 4% uncertainty is the fit covariance only and does not include this calibration error. Please provide a quantitative estimate of this systematic bias, for example from a Monte Carlo using the measured intensity profile, or restrict th","section":"Eqs. (3)-(8) and Fig. 3"},{"comment":"The validation against release-and-recapture (R&R) thermometry uses R&R only at the two endpoints, t=0 and t=1 ms. The 'interpolated R&R' curve is generated by assuming T(t) = T_init + alpha * integral Nsc(t') dt' with alpha chosen to force T(1 ms) = T_end. Thus the mid-range agreement visible in Fig. 4 is not an independent check of the visibility thermometer; it is a consistency check with a model whose free parameter is fixed by the endpoints. The same pump-only scattering data are used in both the heating interpolation and in fixing the amplitude prefactor for the visibility model. This does not make the estimator circular, but it does mean the reported 4% uncertainty does not yet establish absolute accuracy at intermediate temperatures. Please state this limitation explicitly and, if possible, add an intermediate-temperature calibration.","section":"Heating model, Eqs. (13)-(14) and Fig. 4"}],"minor_comments":[{"comment":"The fitted decay rate is given as xi = 0.51/s, but the time axis is in ms and the data visibly decay on a 1-ms scale. This is presumably 0.51/ms; please correct the unit.","section":"Eq. (13)"},{"comment":"Axis label 'Colleted counts' -> 'Collected counts'.","section":"Fig. 7"},{"comment":"The name 'Goncalves' is spelled inconsistently: 'Goncalves' in the abstract and reference [19], but 'Gonçalves' in the acknowledgments. Please standardize.","section":"References and acknowledgments"},{"comment":"The abstract claims 'time-resolved' thermometry with '200 us time resolution.' This is the bin width of an ensemble average over 80 exposures per atom and 1200 atoms, not a single-shot real-time measurement. The text later notes the fit becomes unstable at low photon numbers and above ~70 uK; this limitation should be stated in the abstract or at least in the conclusions.","section":"Abstract and Sec. IV"},{"comment":"The statement 'available from the corresponding author upon reasonable request' is weaker than current best practice; consider depositing the processed histograms and fitting code in a public repository to support reproducibility.","section":"Data availability"}],"recommendation":"major_revision","confidential_remarks":"The central idea is promising and the experiment appears carefully done, but the dimensional inconsistencies in the printed formulas and the unquantified systematic bias in the thermometer are load-bearing for the main claim. I would encourage the editors to request a revision rather than reject, as the issues are addressable within the scope of the paper. The citation list is appropriate and the work is within the scope of the journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Alice,\n\nThe useful thing to know about this paper: it's a real experimental demonstration of an effect that was only predicted before—single-atom resonance fluorescence interfering with an external coherent probe to amplify or deamplify the probe. The data are solid: 1200 atoms, 80 ms exposure per atom, and clean histograms that fit their arcsine-plus-noise model. The thermometry application is plausible: visibility drops as the atom heats, and they resolve 200 µs bins with ~4% precision from the fit covariance. That part deserves credit.\n\nWhat's new here is not the physics principle. Goncalves et al. already calculated the interference; Slodička et al. did interferometric thermometry via self-interference. The new content is the concrete realization with an external probe and the statistical treatment of the photon-count distributions. That is a legitimate contribution.\n\nNow the soft spots, in order of how much they matter.\n\nFirst, the printed low-saturation density-matrix formulas are dimensionally wrong: rho_ee = |Omega|^2/Gamma_0 has units of rate, not a population, and rho_eg = i Omega/Gamma_0 has the same issue. From the subsequent expressions, they clearly used rho_ee ~ |Omega|^2/Gamma^2 and rho_eg ~ i Omega/Gamma, so Eq. 2 is fine, but a published paper cannot have those formulas as written.\n\nSecond, the temperature estimator in Eq. 8 assumes the visibility depends only on the Gaussian phase variance sigma_phi^2. In reality, the scattering amplitude Nsc(r) varies with position through collection efficiency and light shifts, and the trap is not perfectly harmonic. The paper even says the mean count shifts to lower values as the atom heats—that is exactly the amplitude-position correlation. If these correlations are present, visibility is reduced at all temperatures and Eq. 8 will infer a biased T. The comparison to R&R only anchors the 18 µK and 133 µK endpoints with a smooth interpolation in between, so it cannot catch a mid-range systematic. The stress-test note has this right.\n\nThird, the claimed 4% uncertainty is the fit covariance only. It does not include the calibration systematics above. So the headline number is optimistic.\n\nNone of these are fatal. The central interference observation is robust and the model captures the statistics well. But the temperature calibration needs more care if the method is to be trusted at the few-percent level.\n\nWho should read it: anyone working with single neutral atoms or ions needing non-destructive thermometry, and people interested in single-emitter coherent-beam interference. It deserves a serious referee; the correctable issues should be fixed in revision.","headline":"Real demonstration of single-atom interference with an external coherent beam, with a useful thermometry spin-off; the calibration model and a dimension typo need fixing before I'd trust the 4% number.","tokens_in":10816,"tokens_out":4154,"would_cite":true,"duration_ms":44769,"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":"A single 87Rb atom, fluorescing into a weak coherent beam, imprints its thermal motion on the interference visibility, letting photon-count histograms serve as a non-destructive, time-resolved thermometer.","keywords":["resonance fluorescence","single trapped atom","interference visibility","non-destructive thermometry","rubidium-87","photon-counting statistics","optical dipole trap","time-resolved temperature measurement"],"falsifier":"Measure the atom's position distribution directly via fluorescence imaging at ~200 us resolution while simultaneously recording the visibility histograms, and check whether the position variance actually equals k_B T/(m omega^2) at every time bin; alternatively, interleave the visibility readout with an independent thermometer such as resolved-sideband spectroscopy at several temperatures inside 18-133 uK and look for a systematic divergence at the high end, where trap anharmonicity is expected to break the Gaussian-phase assumption.","tokens_in":10039,"feed_emoji":"⚛️","tokens_out":5937,"duration_ms":65897,"temperature":0.7,"pith_summary":"The paper shows that a single trapped rubidium atom can act as its own thermometer: resonance fluorescence from the atom interferes with a weak coherent probe beam, and the visibility of that interference falls as the atom heats up because thermal motion washes out the phase between the two fields. The authors build a photon-counting model of this effect, fit histograms of detected photons, and convert the fitted visibility into a temperature estimate. Using 1200 atoms with 80 ms of illumination per atom, they report 4% relative temperature uncertainty at about 30 microkelvin with roughly 200 microsecond time resolution, all without destroying the atom. This matches release-and-recapture thermometry while being non-destructive and much faster.","feed_headline":"A single atom's glow gauges its temperature to 4%","feed_subtitle":"Non-destructive thermometry with ~200-microsecond resolution, demonstrated on one trapped rubidium atom.","key_machinery":"The load-bearing identity is the visibility-temperature relation: V = 2 sqrt(Nsc Npr)/(Nsc + Npr) * exp(-sigma_phi^2/2), with sigma_phi^2 = 2 k_B T/(m omega^2), where Nsc and Npr are the collected counts from scattered and probe light, and omega is the trap frequency. This is paired with a photon-count distribution formed by convolving the arcsine distribution of the interference term with Poisson counting noise and a Gaussian noise PDF, which lets the authors extract the visibility from histograms and invert it through Eq. 8 to obtain a temperature estimate.","core_discovery":"In a far-off-resonance optical dipole trap, a single 87Rb atom is driven by a strong pump and a weak coherent probe at the same frequency. The collected photon flux in the probe mode follows Ncoll = Nsc + Npr - 2 sqrt(Nsc Npr) cos(phi), so the atom can either amplify or deamplify the coherent beam depending on the relative phase between scattered and probe light. The paper confirms this predicted first-order interference and shows that, after averaging over atomic motion, the visibility decays as exp(-sigma_phi^2/2), where sigma_phi^2 = 2 k_B T/(m omega^2) is the variance of the phase imprinted by the atom's center-of-mass position. Fitting the measured photon-count histograms with a distrib","pith_inferences":["The authors do not go this far, but the same signal could be phase-locked and turned into a continuous, real-time temperature servo, making microsecond-timescale feedback cooling of single atoms practical.","A variant in which the probe is introduced after collection, via a weak beamsplitter and a coherent state, would extend the method to emitters coupled to only one traveling mode; the paper mentions the idea in passing, but its noise floor and practical limits are left open.","Because the measurement is non-destructive, repeated readings on the same atom could track heating dynamics shot by shot rather than averaging over many atoms, at the cost of a more complex statistical reconstruction.","A visible signature of trap anharmonicity would be a systematic drift of the inferred temperature with probe intensity or time-bin position at fixed physical temperature, which would test the Gaussian phase-variance assumption directly."],"forward_implications":["Trapped-atom experiments can monitor center-of-mass temperature without losing the atom, removing the need to reload and recool for every thermometry point.","The demonstrated 4% uncertainty at ~30 uK with ~200 us time resolution improves on the ~10% destructive release-and-recapture benchmark used in the same setup.","The interference signal also offers a path-stabilization handle: because the observed flux depends on the pump-probe phase, the same signal could lock optical path lengths to a single emitter.","At high probe flux the configuration approaches homodyne detection of resonance fluorescence, potentially opening the signal to studies of quantum dynamics.","The method should transfer to other trapped neutral atoms and ions, since it only requires a resonant two-level transition and a coherent probe beam."],"supporting_citations":[{"why":"Supplies the predicted interference scenario between resonance fluorescence and a coherent probe that the experiment confirms and extends to fluctuating conditions.","marker":"[19]"},{"why":"Provides the single-atom loading, cooling, and release-and-recapture thermometry implementation used to benchmark the new method and to interpolate temperature during exposure.","marker":"[21]"},{"why":"Establishes release-and-recapture as the destructive thermometry baseline against which the non-destructive visibility thermometer is compared.","marker":"[11]"},{"why":"Gives the quantum-jump-spectroscopy measurement of the FORT waist, fixing the trap frequency used in the temperature estimator of Eq. 8.","marker":"[30]"}],"fun_headline_variants":["Single atom's glow measures temperature to 4%","One atom's light interference reads its temperature","Laser trick turns single atom into thermometer","Photon interference yields fast atom thermometry","Non-destructive temperature read from one atom"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The temperature estimate assumes the atom's position follows a Boltzmann distribution in a harmonic, Y-Z symmetric trap, so that the phase fluctuations are Gaussian with variance 2 k_B T/(m omega^2); if the trap is anharmonic or the position distribution becomes non-thermal while the atom heats, the inferred temperature is biased.","fun_headline_variants_meta":{"raw":{"variants":["Single atom's glow measures temperature to 4%","One atom's light interference reads its temperature","Laser trick turns single atom into thermometer","Photon interference yields fast atom thermometry","Non-destructive temperature read from one atom"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000454,"raw_usage":{"total_tokens":2114,"prompt_tokens":733,"completion_tokens":1381,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":477,"completion_tokens_details":{"reasoning_tokens":1313}},"tokens_in":477,"tokens_out":1381,"duration_ms":12199,"temperature":1.0,"reasoning_tokens":1313,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T04:21:45.782216+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the atom's position distribution directly via fluorescence imaging at ~200 us resolution while simultaneously recording the visibility histograms, and check whether the position variance actually equals k_B T/(m omega^2) at every time bin; alternatively, interleave the visibility readout with an independent thermometer such as resolved-sideband spectroscopy at several temperatures inside 18-133 uK and look for a systematic divergence at the high end, where trap anharmonicity is expected to break the Gaussian-phase assumption.","supporting_citations":[{"cited_title":"Un- conventional quantum correlations of light emitted by a single atom in free space,","cited_arxiv_id":null,"evidence_quote":"Supplies the predicted interference scenario between resonance fluorescence and a coherent probe that the experiment confirms and extends to fluctuating conditions."},{"cited_title":"Energy distribution and cooling of a single atom in an optical tweezer,","cited_arxiv_id":null,"evidence_quote":"Establishes release-and-recapture as the destructive thermometry baseline against which the non-destructive visibility thermometer is compared."}],"review_version":2}