{"id":"cd9904f5-edb8-4a8a-b1fb-754159e68d83","arxiv_id":"2509.20005","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Droplet oscillation frequency matches passive droplet resonance even with 1 to 6 mesoscale swimmers confined inside, while crowding reduces wobble amplitude via a new scaling law interpreted as slower swimming in 3D.","lead":"Droplets holding one to six swimming Artemia nauplii wobble at the same resonant frequency as plain water droplets, regardless of how many swimmers are inside. The wobble amplitude follows a collision-based scaling law and reveals that crowding slows the swimmers in 3D.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Crowding-speed reduction is inferred from Kc = A/At, not measured; any N-dependent error in Eq. (1) (e.g. sublinear amplitude superposition) appears as Uc/U0 ~ N*^-0.7.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing point: Uc/U0 is inferred from Kc = A/At rather than measured. I agree with that assessment. The frequency collapse onto the CK model (Fig. 2c) and the uncrowded parity plot (Fig. 3b) are solid and give the paper genuine value, but the headline crowding-speed exponent is not independently supported. Because the concern does not overturn the conditional verdict already given, I recommend keeping the verdict conditional: the paper should be published only with the crowding-speed inference clearly labeled as model-dependent, or better, with a direct speed measurement or a validation of the N-scaling in Eq. (1). No ad hominem is intended; the issue is structural in the inference, not in the authors' care.","tokens_in":9417,"tokens_out":4752,"duration_ms":41435,"concrete_test":"Directly track Artemia inside the same 3D droplet geometry (e.g. high-speed volumetric imaging or optical coherence tomography) for N = 1, 2, 4, 6 at fixed droplet volume and swimmer length, and compare the measured Uc/U0 against the values inferred in Fig. 3c. If the directly measured speed ratio does not follow N*^-0.7, the inferred crowding-speed law is an artifact of Eq. (1)'s assumptions. A cheaper complementary check using the existing Zenodo data: regress log A vs log N for fixed V and L; if the slope is significantly below 1, the linear superposition At ~ P A0 N is invalid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim Uc/U0 ~ N*^-0.7 (Fig. 3c) rests on the definition Kc = A/At, introduced in the paragraph defining Kc. There, At is computed from Eq. (1), which assumes At ~ P A0 N with P = 3(L/R) - 3(L/R)^2 + (L/R)^3 and A0 ~ Ua Va/(4 V f1). Every N-dependent discrepancy between the measured amplitude A and Eq. (1) is therefore forced into Kc and interpreted as a speed reduction Uc = Kc Ua. This is internally consistent but circular unless alternative N-dependent effects are excluded. At least three such effects could produce the same downward trend in A/At without any change in swimmer speed: (i) if collision impacts are uncorrelated in time, the steady-state RMS oscillation amplitude should scale roughly as N^{1/2}, not N, making Kc fall as N^{-1/2} purely from the wrong superposition rule; (ii) crowding may change the effective collision probability P by altering swimmer trajectories and interface avoidance, so P itself becomes N-dependent; (iii) synchronized impacts, which the paper itself invokes through the f2 envelope in Fig. 2b, would change the phase relation and momentum-transfer per impact. The paper acknowledges the inference ('we infer reductions in swimmer velocity') but does not falsify these alternatives. The uncrowded parity plot in Fig. 3b validates only the prefactor of Eq. (1) at low N; it cannot validate the N-scaling because the N=1 anchor is used to define U0. Thus the crowding-speed exponent is load-bearing and currently untested.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports experiments on nearly spherical water droplets on superhydrophobic substrates containing 1–6 living mesoscale swimmers (Artemia nauplii, with a few copepod tests). Droplet oscillations are measured via micropipette deflection. The authors find that the dominant oscillation frequency f1 follows the classical Celestini–Kofman resonance scaling f1 ~ V^-0.5 independent of swimmer number, size, and species. For the oscillation amplitude, they propose a scaling model, Eq. (1), in which the amplitude grows with the number of swimmers N, the collision probability P, and a per-swimmer amplitude A0 set by momentum conservation. In the uncrowded regime they report agreement with a fitted prefactor of 1.8. Deviations of the measured amplitude A from the predicted At are then attributed entirely to crowding-induced swimmer speed reduction through Kc = A/At, yielding an empirical power law Uc/U0 ~ N*^-0.7, compared with ~N*^-0.4 for 2D robotic fish. The paper concludes that crowding in 3D reduces mesoswimmer speed more strongly than in 2D.","tokens_in":1752,"tokens_out":1770,"duration_ms":31372,"significance":"If the central inference is valid, this is a valuable experimental contribution: it provides one of the few measurements of crowding effects on mesoscale swimmers in a three-dimensional, near-spherical confinement, and it introduces a non-invasive way to probe internal swimmer dynamics through droplet oscillation amplitudes. The frequency result is clean and appears robust across species and swimmer sizes. The paper also makes its data available on Zenodo, which is commendable. However, the main quantitative claim—the Uc/U0 ~ N*^-0.7 crowding law in Fig. 3c—is not a direct measurement of swimmer speed but is inferred from the ratio A/At. Because any N-dependent error in the amplitude model is forced into Kc, this claim is currently underdetermined. The paper's value would be substantially strengthened by independent speed measurements or by explicit falsification of alternative explanations for the amplitude deficit.","major_comments":[{"comment":"The central crowding result Uc/U0 = (A/At)(Ua/U0) is circular in construction: Kc is defined as A/At, so the measured amplitude deviation from Eq. (1) is, by definition, interpreted as a swimmer speed reduction. Every N-dependent discrepancy between A and At is then forced into Uc. The uncrowded parity plot (Fig. 3b) validates the prefactor of Eq. (1) only at low N and cannot validate the N-scaling, because the N=1 point is used to normalize U0. Alternative N-dependent effects with no speed change—such as sub-linear superposition of uncorrelated impacts (RMS amplitude ~ N^1/2 instead of N), an N-dependent collision probability P, or the synchronized impacts invoked in the f2 envelope—could produce the same downward trend in A/At. The paper acknowledges the inference but does not rule out these alternatives. Please provide independent speed measurements (e.g., direct tracking of swimmers","section":"Paragraph defining Kc; Fig. 3c"},{"comment":"The amplitude model is calibrated with a fitted prefactor of 1.8 (A = 1.8 At) and no uncertainty or error bars are reported on the parity plot or on the fitted exponent in Fig. 3c. Given that Eq. (1) involves several order-unity assumptions (Ud ~ 4 A0 f1, linear superposition At ~ P A0 N, ellipsoidal swimmer volume, P = 3(L/R) - 3(L/R)^2 + (L/R)^3), the prefactor 1.8 could absorb missing physics rather than represent genuine agreement. More importantly, the crowding exponent -0.7 is obtained from a fit whose range and statistical quality are not documented. Please report the number of data points, confidence intervals, and a sensitivity analysis with respect to the chosen cutoff values L* < 0.4 and N* < 0.03.","section":"Eq. (1) and Fig. 3b"},{"comment":"The definition of unconfined and uncrowded via the hand-picked cutoffs L* < 0.4 and N* < 0.03 is not justified, and no sensitivity analysis is provided. The crowding exponent in Fig. 3c depends on which points are included in the normalization U0 and on where the power-law fit begins and ends. With only N = 1–6 swimmers and a narrow range of dimensionless densities, the reported power-law exponents (-0.7 vs -0.4 for the fish data) may not be robust. Please show the fit with confidence bounds and test how the exponent changes with reasonable variations of the cutoffs.","section":"Regime cutoffs L* < 0.4 and N* < 0.03"}],"minor_comments":[{"comment":"Typos: continuos should be continuous; Copeopods should be Copepods; the Bond number definition says Bo = rho_d g R^2 / gamma = 0, but the dimensionless Bond number has gR^2, not R^2 alone; please clarify.","section":"General"},{"comment":"The estimate Ud ~ 4 A0 f1 is introduced without derivation or justification. A sentence explaining the factor 4 and the characteristic time 1/(4 f1) is needed.","section":"Eq. (1) derivation"},{"comment":"The definition of U0 = Kc0 Ua, where Kc0 is the slope of N=1 and L*<0.4 data in Fig. 3b, is confusing because Kc0 already normalizes the amplitude model. Please state explicitly how Kc0 is computed and how its uncertainty propagates to Uc/U0.","section":"Fig. 3c normalization"},{"comment":"The inline reference to See Supplementary Material at [URL] contains a placeholder; if the paper is to be considered for publication, the actual DOI or link must be provided.","section":"Supplementary material reference"},{"comment":"The envelope frequency f2 is mentioned but not analyzed. Since the paper invokes synchronized impacts to explain f2, and synchronization is one of the alternative mechanisms that could affect A/At, a brief discussion or a pointer to future work would help the reader assess the robustness of the crowding interpretation.","section":"Fig. 2 and f2"}],"recommendation":"major_revision","confidential_remarks":"The frequency collapse and the amplitude scaling in the uncrowded regime are interesting and likely publishable. The main risk is the crowding-speed exponent: the inference Uc/U0 = A/At is not a measurement, and the manuscript currently does not provide independent evidence that speed reduction is the only cause of the amplitude deficit. If the authors can add direct speed measurements, or at least clearly reframe the result as an amplitude-based effective speed subject to model assumptions, the paper could become suitable. I would advise against acceptance until the circularity is addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The frequency result is clean and the amplitude scaling law is physically sensible, but the headline crowding-velocity exponent is an inference built on a circular definition. Treat that part as a hypothesis, not a measurement.\n\nWhat is genuinely new: this is the first experimental look at living droplets at swimmer-size comparable to droplet radius, and the data show that the resonant frequency still follows the classical CK model regardless of swimmer number, size, or species. That collapse is a solid, reproducible observation. The amplitude model — At ~ Ua Va N /(4V f1) times the peripheral-volume factor — is a simple momentum-conservation argument, and it works for the uncrowded, unconfined cases within a fitted prefactor of 1.8. The setup is careful, with micropipette calibration, control experiments without the pipette, and the data are shared on Zenodo.\n\nThe soft spot is the crowding analysis. The factor Kc is defined as A/At, and then interpreted as a velocity reduction Uc = Kc Ua. Any N-dependent error in Eq. (1) — for example, if the amplitude superposition is sublinear (uncorrelated impacts would give √N rather than N), if the collision probability itself changes with crowding, or if the envelope mode f2 reflects phase synchronization — will appear as a “speed reduction.” The paper acknowledges that this is an inference, but it does not test these alternatives. The uncrowded parity plot validates the prefactor only at low N, and the N=1 point is used to define U0, so it cannot validate the N-scaling. The exponent -0.7 is therefore not falsifiable by the present data. The comparison with 2D robotic fish uses a different density metric and different Reynolds number, so the claimed 3D-vs-2D difference is suggestive at best.\n\nNone of this kills the paper. The frequency result is worth having, and the amplitude scaling law is a useful starting point for thinking about internally actuated droplets. What needs work is the crowding claim. A direct measurement of swimmer speed inside the droplet, or at least a test of the linear superposition assumption, would turn the -0.7 exponent from an artifact of the model into a real result. As it stands, I would not cite the exponent, but I would cite the experimental system and the frequency/amplitude results.\n\nSend it to peer review — it is novel and careful enough to deserve referee time — but the reviewers should be asked to focus on validating the crowding inference. This would make a good reading-group paper, if only to discuss model circularity.","headline":"Frequency result is clean, amplitude model plausible, but the crowding-velocity exponent is an inference built on a circular definition — treat as hypothesis, not measurement.","tokens_in":10358,"tokens_out":5390,"would_cite":true,"duration_ms":96953,"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":"Inside a droplet, swimming organisms cannot change its resonant frequency; they only change how hard it shakes, and crowding slows them as a -0.7 power law.","keywords":["living droplets","mesoscale swimmers","droplet resonance","crowding","Artemia nauplii","superhydrophobic substrates","oscillation amplitude scaling","3D active matter"],"falsifier":"Track the actual 3D swimming paths of individual Artemia inside droplets as N increases; if their directly measured mean speed does not fall as N*^-0.7 while the droplet amplitude still drops, then the Kc inference is wrong. Alternatively, run the same droplet experiment with inert particles of identical size and density: if amplitude falls with density without any living slowdown, the collision model itself carries the deficit.","tokens_in":9383,"feed_emoji":"💧","tokens_out":6778,"duration_ms":53023,"temperature":0.7,"pith_summary":"The paper tries to establish how internally generated swimming activity changes the mechanics of a water droplet. Using droplets on a superhydrophobic surface that contain one to six brine shrimp larvae, it shows that the droplet's oscillation frequency is the same as the classical resonance of an inert droplet, regardless of how many swimmers are inside or how they swim. The oscillation amplitude, however, is set by the probability of swimmers colliding with the interface and by the momentum each collision transfers, captured in a new scaling law. In the dilute, unconfined regime the law predicts measured amplitudes; as swimmers crowd the droplet, the amplitude deficit implies their swimming speed falls as the -0.7 power of dimensionless density. If true, this turns a droplet into a non-invasive readout of crowded-swimmer kinematics in 3D.","feed_headline":"Crowding slows swimming organisms inside droplets by a power law","feed_subtitle":"A droplet's shake frequency ignores its swimmers; its amplitude reveals their speed drops as crowding rises.","key_machinery":"The load-bearing object is Eq. (1), a scaling law for droplet amplitude built from three steps: a geometric collision probability P = 3L/R - 3(L/R)^2 + (L/R)^3, equal to the peripheral volume fraction within one swimmer length of the interface; momentum conservation during a single swimmer–interface impact, which gives the per-swimmer amplitude A0 ~ Ua Va/(4 V f1); and a first-order superposition At ~ P A0 N for N swimmers. The paper then defines a crowding–confinement factor Kc = A/At, where A is the measured RMS amplitude, and interprets Kc as the ratio of crowded to free swimming speed. This factor converts a passive droplet measurement into an inference about swimmer kinematics.","core_discovery":"The central discovery is a clean separation of time scales: the living droplet's frequency response is dictated by the droplet itself and is insensitive to the swimmer activity, while the amplitude is a direct measure of that activity. For one to six Artemia nauplii in droplets 600–1750 µm in radius (L/R = 0.30–0.57), the measured fundamental frequency follows the classical sessile-droplet resonance scaling f1 ~ V^-0.5. The authors derive an amplitude law, At ~ (Ua Va N)/(4 V f1)[3(L/R) - 3(L/R)^2 + (L/R)^3], from collision probability and momentum conservation, and find agreement with experiment (A = 1.8 At) in the uncrowded regime. They then attribute the amplitude shortfall at higher dens","pith_inferences":["If the Kc identification is correct, droplet amplitude becomes a general speedometer for crowded 3D swimmers; a direct test would be to track individual larvae in a droplet with volumetric microscopy and compare measured speeds with Uc from Fig. 3c.","The -0.7 versus -0.4 exponent gap is likely shaped by the 3D collision geometry at L ~ R; varying L/R at fixed density should interpolate the exponent toward the 2D value—a prediction the paper does not make.","The low-frequency envelope mode f2 may encode synchronized impacts; if so, controlling swimmer number and phase could tune droplet actuation, but the paper explicitly does not investigate this mode.","Replacing living swimmers with inert particles of the same size and density would isolate the collision and momentum-transfer parts of Eq. (1); if the amplitude deficit persists, the speed reduction interpretation would need revision."],"forward_implications":["The droplet's resonant frequency is an invariant property of the droplet itself, so swimmer-laden droplets can be used as resonators whose frequency is set by volume, surface tension, and contact angle rather than by swimmer activity.","Eq. (1) gives a predictive formula for oscillation amplitude from swimmer size, number, droplet volume, and unconfined swim speed in the dilute regime.","Crowding slows mesoscale swimmers in 3D more steeply than previously observed in 2D (exponent -0.7 vs -0.4), indicating dimensionality changes the collective slowdown mechanism.","The micropipette readout is sensitive enough to detect collective swimmer effects without imaging individual organisms, opening a non-invasive probe for dense active matter in 3D.","The platform could be scaled to organisms too small or too fast for direct video tracking, extending resonance-based activity sensing to microbial swimmers."],"fun_headline_variants":["Crowding slows swimmers inside droplets by a power law","Droplet shake amplitude reveals swimmer crowding slowdown","Living droplets: frequency constant, amplitude tracks swarm speed","Power-law slowdown of mesoscale swimmers in crowded droplets","Droplet oscillations measure swimmer speed drop with crowding"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The speed-reduction curve rests on the assumption that every deviation between the measured amplitude and Eq. (1) is caused solely by slower swimming; if crowding also changes collision probability, the linear superposition of single-swimmer amplitudes, or the momentum transfer per collision, then the inferred N*^-0.7 scaling is an artifact of the model rather than a measured speed.","fun_headline_variants_meta":{"raw":{"variants":["Crowding slows swimmers inside droplets by a power law","Droplet shake amplitude reveals swimmer crowding slowdown","Living droplets: frequency constant, amplitude tracks swarm speed","Power-law slowdown of mesoscale swimmers in crowded droplets","Droplet oscillations measure swimmer speed drop with crowding"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000361,"raw_usage":{"total_tokens":1731,"prompt_tokens":636,"completion_tokens":1095,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":380,"completion_tokens_details":{"reasoning_tokens":1017}},"tokens_in":380,"tokens_out":1095,"duration_ms":12601,"temperature":1.0,"reasoning_tokens":1017,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T15:13:51.006154+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Track the actual 3D swimming paths of individual Artemia inside droplets as N increases; if their directly measured mean speed does not fall as N*^-0.7 while the droplet amplitude still drops, then the Kc inference is wrong. Alternatively, run the same droplet experiment with inert particles of identical size and density: if amplitude falls with density without any living slowdown, the collision model itself carries the deficit.","supporting_citations":[],"review_version":1}