{"id":"7145350f-aff0-411e-8410-43a8da64f466","arxiv_id":"2608.13346","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"An automatic algorithm in the CLASS Boltzmann code detects the onset of rapid oscillations in scalar-field dark matter and enables accurate averaging even with dark matter-dark energy interactions; updated QCDM constraints still favor ΛCDM.","lead":"This paper proposes a way for cosmology software to automatically detect when a hypothetical dark matter particle starts oscillating rapidly, so the equations can be switched to a faster averaged form without guessing. The authors test it on an interacting dark matter-dark energy model and use the latest data to update its constraints, finding the model fits the data about as well as the standard ΛCDM cosmology.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Automatic detection is validated only in the non-interacting limit and at one benchmark; fixed thresholds may fail across the QCDM prior, biasing the reported constraints.","rationale":"The reader's weakest-assumption and this stress-test converge: the linchpin of the paper is the automatic detector, and its fixed thresholds are not justified across the QCDM prior. The WKB derivation in Sec. 4 and the averaging construction in Sec. 5 are internally plausible and give the method genuine independent support: the zero-average-pressure condition follows from the separation of scales, and the T(θ) interpolation is a reasonable smooth handoff. The MCMC pipeline and data choices are standard, and the reported constraints are internally consistent. However, the automatic detection is the gate through which all results pass. If the detector triggers too early or too late, the averaged background and perturbations are wrong, and the posterior constraints in Table 3, the Bayes factors, and the claimed comparison to LCDM all inherit that bias. The paper offers no sensitivity study and no comparison to exact evolution for any interacting parameter point, and its analytic motivation is explicitly non-interacting. That does not prove the method is wrong, but it makes the central numerical claim unverified as it stands. A parameter-space scan with an exact-reference comparison is the direct check that would settle the issue, which is precisely the condition of the reader's CONDITIONAL verdict. The absence of a repository link or hash is an additional reproducibility concern, but it is secondary to the physical question of whether the thresholds generalize.","tokens_in":19333,"tokens_out":12079,"duration_ms":120292,"concrete_test":"Run the public CLASS fork in automatic mode on a grid spanning the sampled priors, e.g. lambda=1e-2,1e3,1e6,1e9 Mpc^-2 and phi_ini=0.05,0.5,0.95 M_Pl, holding other parameters near the Table 3 best fit. For each point, record the detected switching redshift z_det. Independently determine the true onset z_true by integrating the full Klein-Gordon system with a high-accuracy stiff solver until H/m_eff drops below a conservatively small value, and compare the resulting C_l^TT and P(k) between the automatic run and this exact reference. If z_det differs from z_true by more than one local oscillation period, or if any C_l or P(k) changes by more than 0.1% for any grid point, the fixed thresholds are not robust across the QCDM parameter space.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on the automatic detector identifying the true onset of oscillations across the entire QCDM parameter space. The only analytic justification, Eq. (6.4), is derived from the non-interacting equations (6.1)-(6.2), and the paper explicitly states this is 'for illustrative purposes.' In QCDM, theta and y are defined using V1 only (Eqs. (3.22)-(3.24)), while the KG equations (3.14)-(3.15) contain Vint,chi and Vint,phi; the evolution equations for theta and y therefore acquire interaction source terms that are not given in the paper (deferred to ref. [32]). The fixed trigger f_R=0.70 and the confirmation thresholds in Table 1 are calibrated, at most, on this non-interacting dynamics, and the single benchmark lambda=1e5 Mpc^-2 in Fig. 1 does not cover the sampled prior lambda in [1e-2,1e9] Mpc^-2 in Table 2. If, for large lambda, m_eff^2=m_chi^2+lambda*phi^2 changes appreciably while phi rolls, the plateau drop in r=y/theta may be shallower than 30%, occur at a different theta, or be accompanied by transients that defeat the running-maximum logic; the sign-change test only confirms that w_chi oscillates, not that the chosen theta*=max(theta_drop+5,6) coincides with the true onset. A premature or delayed averaging switch would bias the background and perturbation spectra and hence every constraint in Table 3.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents an automatic method to detect the onset of fast oscillations of a scalar-field dark matter component and to switch from the exact Klein-Gordon evolution to an averaged fluid description when interactions with a dark-energy field are present. The detector monitors the ratio r=y/theta, flags a 30% drop from its running maximum, and requires a separate confirmation from sign changes of the dark-matter equation of state; the transition phase is then set to theta* = max(theta_drop + 5, 6). The averaging uses a tanh transition function T(theta) and replaces interaction terms by their oscillation averages so that <p_chi>=0 and <delta p_chi>=0 in the oscillating regime. The scheme is implemented in CLASS, demonstrated on the QCDM model at lambda=10^5 Mpc^-2, and used in MCMC analyses with Planck, ACT, SPT-3G, DESI DR2, and Pantheon+ data. The resulting constraints favor LCDM, with Bayes factors ln B between -0.569 and -2.189 for the data combinations considered.","tokens_in":19715,"tokens_out":7980,"duration_ms":78929,"significance":"If the automatic detector and the interacting averaging scheme are reliable, the method would be a practical improvement for scanning scalar-field dark matter models, since it removes the need to guess a transition threshold by hand. The consistency conditions in Secs. 4 and 5 are physically sensible, and the public CLASS implementation together with the MCMC pipeline are useful contributions. The main significance is methodological; the updated cosmological constraints are a secondary application. However, the current validation rests on a non-interacting analytic example and a single interacting benchmark, which is insufficient to establish the central claim that the detector and averaging work across the QCDM parameter space.","major_comments":[{"comment":"The analytic derivation of the ratio-drop criterion is performed only for the non-interacting equations (6.1)-(6.2), and the text explicitly states that this case is used \"for illustrative purposes.\" In QCDM, theta and y are defined using V1 only (Eqs. (3.22)-(3.24)), while the field equations (3.14)-(3.15) contain V_int,chi and V_int,phi; the evolution equations for theta and y therefore acquire interaction source terms that are not presented. The fixed detector parameters in Table 1 (f_R=0.70, theta_min=6, etc.) are then applied over the entire prior log lambda in [-2,9] of Table 2, but the only interacting validation is the single benchmark lambda=10^5 Mpc^-2 in Figs. 1-5. This is a load-bearing gap because a premature or delayed switch changes the predicted background and perturbation evolution and hence the constraints in Table 3. Please provide the interacting theta and y equations, and validate the detector on a grid of lambda (and m_chi) values by comparing the automatic theta* with the true onset from exact integration, or by demonstrating that moderate changes in theta* do not alter the predicted spectra beyond the claimed accuracy.","section":"Sec. 6, Eq. (6.4)"},{"comment":"The averaging prescription for perturbations is asserted rather than derived. The paper defines delta V_int and delta p-tilde_chi so that <delta p_chi> -> 0 as T -> 0, but it does not show that the resulting averaged perturbation equations correctly reproduce the exact Klein-Gordon perturbation evolution at the transition. The only numerical evidence, Figs. 3-5, is a QCDM-versus-LCDM comparison at lambda=10^5 Mpc^-2 and says nothing about the error introduced by the averaging switch. Since the stated purpose of the method is to avoid biases at the handoff (cf. refs. [29,30]), please add a direct comparison of the averaged background and perturbation outputs to the exact evolution for at least the benchmark and a few representative points across the prior, with the target accuracy quantified.","section":"Sec. 5, Eqs. (5.13)-(5.15)"},{"comment":"The automatic detector introduces seven user-specified parameters, none of which is derived or tested for sensitivity. The paper's advertised advantage over the manual gamma* is that the user need not supply a threshold, but the algorithm still depends on f_R=0.70, w=5, theta_floor=10^-12, theta_min=6, w_min=0.10, min_sign_changes=2, and theta_pad=5. For large lambda, where m_eff^2=m_chi^2+lambda phi^2 varies appreciably while phi rolls, the r=y/theta plateau drop may be shallower than 30% or occur at a different theta, and the sign-change test only confirms that w_chi oscillates, not that theta* coincides with the true onset. Please provide a sensitivity analysis of the detector parameters and report the distribution of detected theta* (or transition redshift) over the MCMC samples; otherwise the robustness of the Table 3 constraints to the detector choice is not established.","section":"Table 1"}],"minor_comments":[{"comment":"\"tbe perturbations\" should read \"the perturbations.\"","section":"Sec. 3, after Eq. (3.21)"},{"comment":"The caption does not state which manual threshold was used for the lower panels; please add it for reproducibility.","section":"Fig. 2"},{"comment":"Please clarify that n indexes accepted background time points and define the range of n used in the running maximum R_n.","section":"Sec. 6, Eqs. (6.6)-(6.7)"},{"comment":"The entry log(lambda/Mpc^-2) < 3.99 for CMB+DESI+PPS should state whether this is a 68% or 95% upper limit.","section":"Table 3"},{"comment":"\"Prosum\" appears to be a typo for \"Program.\"","section":"Acknowledgments"},{"comment":"Several DOIs, e.g., refs. [32], [45], [46], and [55], have nonstandard formats; please verify them against the publisher records.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The reader's report and my reading agree that the central methodological claim needs additional validation. The paper would be strengthen significantly by adding a validation suite for the automatic detector across the QCDM parameter space, a sensitivity study of the Table 1 parameters, and a direct comparison of averaged versus exact perturbation evolution. I also note that the manuscript relies heavily on ref. [32] for the perturbation equations; the missing derivations should either be included or clearly referenced. The nonstandard DOI strings in several references may indicate production issues that should be checked before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Automatic onset detection via the r=y/theta drop with a sign-change confirmation is a genuinely new algorithmic idea, and the interaction-aware averaging is a real extension beyond non-interacting axion treatments. The validation, though, is thin enough that the numerical claim should be stress-tested before the constraints are taken at face value.\n\nWhat is actually new: earlier methods require a user-supplied threshold (e.g., gamma_* ~ 100 in the polar-variable formalism). Here the code learns the plateau of r, detects a 30% drop, and then confirms oscillations via sign changes in w_chi. I do not see that combination in prior work. The averaging prescription in Sec. 5, which makes the interaction contribution to <p_chi> vanish in the T->0 limit, is also new and looks correct in the limit shown. The paper is honest about the main model comparison: the full data give lnB = -2.189 relative to LCDM, so the data do not prefer QCDM.\n\nThe soft spots, in order of importance. First, the analytic justification of the detection criterion, Eq. (6.4), is derived only for the non-interacting case, and the paper says so explicitly. In QCDM, theta and y use V1 only, while the KG equations contain interaction sources; the interacting evolution equations for theta and y are not given. For large lambda, m_eff^2 = m_chi^2 + lambda phi^2 changes while phi rolls, and the r-drop may be shallower or occur at a different theta than the fixed thresholds assume. The validation for the interacting case is a single benchmark lambda=1e5 Mpc^-2, while the MCMC prior spans lambda in [1e-2,1e9] Mpc^-2. No sensitivity study of the hand-chosen thresholds (f_R=0.70, w=5, padding, etc.) is presented. That is not fatal, but it means the central numerical claim is under-supported. Second, the code is stated to be on Github, but there is no URL or commit hash, so the implementation cannot be checked. Minor, and addressable.\n\nWhere the paper is solid: the discussion of why naive averaging fails with interactions (Sec. 4) is clear, the perturbation construction in Sec. 5 is carefully laid out, and the MCMC setup is standard.\n\nWho this is for: people integrating scalar-field DM into Boltzmann codes, especially for parameter scans with DM-DE interactions. They will want to read it and probably test the detector on their own models. The paper deserves a serious referee, not a desk reject; the missing pieces are addressable in revision. My recommendation is to send it to review with a request for (a) a sensitivity analysis of the thresholds across the sampled lambda range, (b) a code link or hash, and (c) at least one additional benchmark where averaged spectra are compared to exact evolution.","headline":"Real algorithmic novelty in automatic oscillation-onset detection, but the validation is thin enough that the numerical claim should be checked before trusting the derived constraints.","tokens_in":20197,"tokens_out":3661,"would_cite":true,"duration_ms":34454,"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 claims that the onset of rapid oscillations in scalar-field dark matter can be detected automatically from the field dynamics, triggering a consistent oscillation-averaged treatment without a user-supplied threshold, even when…","keywords":["scalar field dark matter","rapid oscillations","automatic onset detection","oscillation averaging","QCDM model","dark matter-dark energy interaction","cosmological constraints","Boltzmann solver"],"falsifier":"Run the CLASS module with the required ratio drop varied from 0.5 to 0.9 on a fixed QCDM benchmark and rerun the MCMC fits: if the recovered $H_0$, $S_8$, or interaction-strength posteriors move by more than their quoted 68% intervals, the automatic onset is not threshold-independent. Alternatively, compare the detected $\\theta_*$ with the time at which the WKB-averaged equation of state $\\langle w_\\chi\\rangle$ first vanishes to 1% accuracy in a non-interacting quadratic model over a grid of masses; a systematic offset would indicate the detector tracks a proxy rather than the true transition.","tokens_in":19134,"feed_emoji":"🌀","tokens_out":11206,"duration_ms":96719,"temperature":0.7,"pith_summary":"This paper claims that the onset of rapid oscillations in a scalar-field dark matter component can be detected automatically from the field's own dynamics, removing the need for a user-supplied threshold that is usually tuned by trial and error. The detector watches the ratio $r = y/\\theta$ of two phase-space variables; a sudden drop in this ratio marks the transition to oscillatory behavior, and the interpretation is independently confirmed by requiring the dark matter equation of state $w_\\chi = -\\cos\\theta$ to change sign repeatedly. The paper further shows that a naive averaging of the equations fails when dark matter interacts with dark energy, leaving a spurious interaction-induced pressure, and presents a modified averaging scheme that keeps the averaged equation of state zero in the oscillating regime. Implemented in the CLASS Boltzmann solver and applied to the interacting QCDM model, the technique enables scans of the parameter space and yields updated constraints from recent BAO, supernova, and CMB data, with Bayesian evidence still favoring the standard cosmological model.","feed_headline":"No more hand-tuning: code detects onset of dark-matter oscillations","feed_subtitle":"The new trigger also handles dark matter–dark energy interactions and updates QCDM fits against current BAO, supernova and CMB data.","key_machinery":"The central object is the ratio $r = y/\\theta$ formed from the polar-angle variables $(\\Omega_\\chi,\\theta,y)$ introduced to recast the Klein-Gordon equation as first-order flow equations. In radiation domination the attractor $r_0 = \\beta + 3\\sin\\theta/\\theta$ drops from 5 to 2 as $\\theta$ grows, so a drop in $r$ marks the onset of oscillations; the detector learns the running maximum $R_n$ and triggers on $r_n \\le 0.70\\,R_n$, then confirms with sign changes of $w_\\chi = -\\cos\\theta$. The averaging is carried by the transition function $T(\\theta) = \\tfrac{1}{2}\\bigl(1 - \\tanh\\bigl((\\theta-\\theta_*)/w\\bigr)\\bigr)$ that interpolates exact and averaged variables, with $\\theta_*$ set automatically from the detected drop plus padding, ensuring $\\langle p_\\chi\\rangle = 0$ in the oscillatory regime even with a dark matter-dark energy interaction.","core_discovery":"The central claim is that the transition to the fast-oscillation regime of a scalar-field dark matter component can be detected automatically with no user-specified onset, and that a modified averaging scheme can then be applied consistently even when the dark matter couples to a quintessence field. The detector tracks the dimensionless ratio $r = y/\\theta$ in the $(\\Omega_\\chi,\\theta,y)$ background variables; at the onset of oscillations $r$ drops from its plateau value (5 in radiation domination) to a lower attractor (2). The code flags a candidate drop when $r$ falls by at least 30% below its running maximum, then requires that the dark matter equation of state $w_\\chi = -\\cos\\theta$ change sign at least twice as an independent confirmation before averaging begins. The averaging uses a transition function $T(\\theta)$ to blend exact and averaged quantities so that the interaction term contributes no spurious pressure, preserving a zero averaged equation of state in the oscillating regime. Applied to the QCDM model, the method reproduces consistent dynamics and yields updated constraints: $H_0 = 68.37^{+0.30}_{-0.26}\\,\\mathrm{km\\,s^{-1}\\,Mpc^{-1}}$ with the full data combination, while the Bayes factor ($\\ln B = -2.189$) still favors $\\Lambda$CDM.","pith_inferences":["The analytic derivation of the ratio drop assumes no interaction, so the robustness of the automatic trigger in strongly coupled regions is an open question; a threshold-sensitivity scan varying $f_R$ and the transition width $w$ could test whether the recovered posteriors drift outside the quoted errors.","Because the detector operates on the field variables themselves rather than a fixed $H/m_\\chi$ criterion, it may transfer to other scalar dark matter potentials (quartic, axionic) with only the sign-change confirmation test needing adaptation.","The manual-mode comparison in the paper shows the automatic trigger fires earlier than the classical cutoff prescription; quantifying how much of the difference in predicted CMB and matter-power spectra comes from the choice of onset versus the improved interaction averaging would make the padding parameter practically motivated rather than heuristic."],"forward_implications":["Large parameter scans and MCMC analyses of scalar-field dark matter no longer need a hand-tuned oscillation-onset threshold; the detector finds the transition separately at each model point.","The averaging method removes the spurious interaction-induced pressure that a naive cutoff would introduce, so the dark matter equation of state correctly returns to zero in the oscillation regime.","QCDM suppresses the late-time matter power spectrum by roughly 30-35% relative to $\\Lambda$CDM while leaving the CMB nearly unchanged, with the growth-rate difference accumulating below the percent level into about a 16% amplitude reduction.","Updated data combinations raise $H_0$ to $68.37 \\pm 0.30\\,\\mathrm{km\\,s^{-1}\\,Mpc^{-1}}$ and lower $S_8$ to $0.8119 \\pm 0.0079$, but the Bayesian evidence for QCDM over $\\Lambda$CDM is negative ($\\ln B = -2.189$ for the full dataset)."],"supporting_citations":[{"why":"Supplies the $(\\Omega_\\chi,\\theta,y)$ variable transformation that turns the Klein-Gordon equation into first-order flow equations.","marker":"[36]"},{"why":"Adapts these variables to scalar-field dark matter and provides the $y$ variable used in the detection ratio.","marker":"[37]"},{"why":"The averaging and cutoff method that the new automatic trigger replaces, along with the $w_\\chi = -\\cos\\theta$ formulation.","marker":"[19]"},{"why":"The Boltzmann solver code in which the detection and averaging scheme is implemented.","marker":"[26]"},{"why":"Derives the background and perturbation evolution equations for the interacting QCDM fields that the new averaging scheme modifies.","marker":"[32]"},{"why":"Defines the QCDM model, including the DM and DE potentials and the interaction strength, whose constraints are updated.","marker":"[25]"},{"why":"Documents how averaging inaccuracies bias precision observables, motivating automated and accurate onset detection.","marker":"[29]"},{"why":"Provides an improved effective-fluid description that sets the accuracy target the new averaging must match.","marker":"[30]"}],"fun_headline_variants":["Auto-trigger for dark-matter oscillations ends manual tuning","Oscillation onset found automatically even for interacting dark matter","CLASS code auto-detects dark-matter oscillation onset, updates QCDM fits","Scalar dark matter oscillations: automatic onset detection, updated constraints"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The detector is trusted to use one fixed set of thresholds (a 30% ratio drop, minimum phase $\\theta_{\\min} = 6$, two sign changes, and related settings) across the whole QCDM parameter space, although the ratio-drop criterion is derived only in the non-interacting limit.","fun_headline_variants_meta":{"raw":{"variants":["Auto-trigger for dark-matter oscillations ends manual tuning","Oscillation onset found automatically even for interacting dark matter","CLASS code auto-detects dark-matter oscillation onset, updates QCDM fits","Scalar dark matter oscillations: automatic onset detection, updated constraints"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000821,"raw_usage":{"total_tokens":3638,"prompt_tokens":1038,"completion_tokens":2600,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":654,"completion_tokens_details":{"reasoning_tokens":2526}},"tokens_in":654,"tokens_out":2600,"duration_ms":16290,"temperature":1.0,"reasoning_tokens":2526,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:14:28.508427+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the CLASS module with the required ratio drop varied from 0.5 to 0.9 on a fixed QCDM benchmark and rerun the MCMC fits: if the recovered $H_0$, $S_8$, or interaction-strength posteriors move by more than their quoted 68% intervals, the automatic onset is not threshold-independent. Alternatively, compare the detected $\\theta_*$ with the time at which the WKB-averaged equation of state $\\langle w_\\chi\\rangle$ first vanishes to 1% accuracy in a non-interacting quadratic model over a grid of masses; a systematic offset would indicate the detector tracks a proxy rather than the true transition.","supporting_citations":[{"cited_title":"Cosmology of axion dark energy in supersymmetric models and constraints on high scale parameters","cited_arxiv_id":"2511.22559","evidence_quote":"Derives the background and perturbation evolution equations for the interacting QCDM fields that the new averaging scheme modifies."}],"review_version":1}