{"id":"58951cc2-47e7-44a7-ac16-a6ba9da22058","arxiv_id":"2504.18464","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Adding ACT DR6 data to DESI DR2 BAO keeps the roughly 3 sigma preference for evolving dark energy and, for the baseline Planck+ACT combination, reports a neutrino mass limit below 0.061 eV.","lead":"This paper combines the Dark Energy Spectroscopic Instrument's new galaxy-distance measurements with the latest sky maps from the Atacama Cosmology Telescope. The combination still favors dark energy that changes over time, and it gives one of the tightest neutrino mass limits from cosmological data.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The robustness claim rests on treating the cut-based Planck+ACT splice as unbiased; an unmodeled inter-survey covariance or relative calibration error could shift the ~3σ dark-energy significance, so the claim is conditional pending a joint treatment.","rationale":"I read the paper as a careful data-combination exercise whose central claim is that the DESI DR2 BAO preference for w0waCDM survives inclusion of ACT DR6. The evidence for this is substantial: the significance is 2.9–3.0σ for the three main CMB combinations without SNe, and the variation across CMB choices is only ~0.5σ. The paper is transparent about its choices, reports the ACT-alone 2.7–3.2σ tension with DESI in ΛCDM, and provides profile-likelihood checks in Appendix B. There is no circularity: the significance comes from fitting external BAO, CMB, and SNe data to w0waCDM versus ΛCDM. The weakest point is not the analysis itself but the assumption that Planck and ACT can be spliced by disjoint multipole cuts without modeling cross-survey covariance or a relative calibration term. Section II and Section VI both acknowledge this limitation, and the sensitivity of Ωch2 and of the neutrino bound to the exact choice of cuts shows the assumption is not innocuous. The reader identified the same concern, and I agree. I do not see grounds to reject: the dark-energy significance is stable enough across the tested variations that a joint treatment is more likely to refine than to erase it. However, the headline 'robust' claim and the tightest neutrino bound should remain conditional until the joint-covariance check is performed. The SN-dependent spread (2.8–4.2σ) is a reporting issue for the abstract rather than a flaw in the central argument.","tokens_in":26202,"tokens_out":5445,"duration_ms":60453,"concrete_test":"Re-run the baseline DESI+PR4+ACT analysis (Table II and Section V) with a joint likelihood that includes cross-survey covariance between Planck PR4 and ACT DR6; at minimum, add one free relative calibration parameter between the two high-ℓ spectra, or construct a joint covariance from the public Planck and ACT likelihood products. Compare the resulting w0waCDM Δχ², significance, and Σmν 95% bound against the cut-based values (Δχ²=−11.9, 3.0σ; Σmν<0.061 eV). If the shifts are ≤0.3σ in significance and ≤0.005 eV in the bound, the robustness claim stands; if the significance drops below 2.5σ or the neutrino bound moves by more than 10%, the paper should present the dark-energy and neutrino results as conditional on the cut-splicing approximation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is that PR4+ACT can be formed by splicing the Planck PR4 and ACT DR6 likelihoods at ℓ_TT=2000 and ℓ_TE,EE=1000 with no cross-survey covariance and no relative calibration term. Section II states this explicitly: the high-ℓ spectra are combined 'without modeling a covariance between the surveys, but rather by applying simple data cuts.' This matters because the w0waCDM preference is driven by the CMB-inferred values of Ωch2, Ωbh2, and ns (Section IV and Appendix A), and these parameters shift with the splice point: the Ωch2 central value changes between P-ACT and PR4+ACT, and the neutrino bound moves from <0.061 eV to <0.077 eV depending on the CMB combination and low-ℓ EE likelihood (Section V). The paper itself concedes in Section VI that 'a joint treatment of Planck and ACT DR6 covariance could eventually coalesce to a consolidated CMB dataset.' With a baseline significance of only 3.0σ (Table II), an unmodeled calibration offset or ℓ-space covariance at the 0.5σ level in Ωch2 could move the headline significance below the conventional 3σ threshold. Since the central conclusion—'the results presented in [16] are robust in light of the new ACT CMB data'—is precisely a claim about the stability of these numbers, the cut-splicing approximation is the least secure link in the argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines DESI DR2 BAO measurements with the ACT DR6 CMB likelihood and several Planck PR3/PR4 likelihoods, spliced at multipole cuts, to test whether the DESI DR2 dark-energy and neutrino-mass conclusions survive the inclusion of ACT. The baseline DESI+PR4+ACT combination prefers w0waCDM over LambdaCDM at about 3.0 sigma without supernovae and up to 4.2 sigma with DESY5 supernovae; across CMB combinations the preference ranges from 2.4 to 3.0 sigma (2.5 to 4.2 sigma with SNe). The same analysis yields sum m_nu < 0.061 eV at 95% CL in LambdaCDM for the baseline combination, compared with < 0.077 eV for the P-ACT combination. The authors conclude that the DESI DR2 results are robust to the inclusion of ACT DR6, while noting that the neutrino-mass constraints are sensitive to the CMB likelihood combination and to the physical prior sum m_nu > 0.","tokens_in":26433,"tokens_out":7118,"duration_ms":73149,"significance":"The paper is a timely and useful cross-check of a high-impact cosmological claim, and it is transparent about the main choices and limitations: several Planck+ACT splice rules are compared, the prior-dependence of the neutrino bound is studied with a profile likelihood in Appendix B, and the dataset variations are clearly tabulated. If the results hold, the 0.061 eV bound is the tightest BAO+CMB neutrino-mass limit reported to date, and the persistence of the w0waCDM preference across CMB combinations strengthens the DESI DR2 evidence. The analysis is not circular: the dark-energy significance comes from fitting external data, and the ACT-based check is a genuine robustness test. The main caveat is that the spliced CMB likelihood contains no cross-survey covariance or relative calibration term, so the headline robustness claim is conditional on that approximation being unbiased.","major_comments":[{"comment":"The baseline PR4+ACT combination is constructed by splicing independent Planck PR4 and ACT DR6 likelihoods at hard multipole cuts, and Section II states that this is done 'without modeling a covariance between the surveys.' All of the headline numbers (the 3.0 sigma w0waCDM preference and the 0.061 eV neutrino bound) come from this spliced likelihood, and the central conclusion that the DESI DR2 results are robust to ACT is a statement about the stability of those numbers. Appendix A and Section V show that the inferred Omega_ch2 and the neutrino upper limit shift with the cut choice (0.061 vs 0.077 eV; a 1.4 sigma vs 2.0 sigma DESI tension), so an unmodeled relative calibration error or inter-survey covariance at the roughly 0.5 sigma level in Omega_ch2 could move the baseline 3.0 sigma significance below the conventional threshold. A quantitative robustness test, for example marginalizing over a relative calibration amplitude or adding an inter-survey covariance term, is needed before the 'robust' conclusion can be regarded as fully load-bearing.","section":"Section II, Table I, Appendix A"},{"comment":"The neutrino-mass limits are prior-dominated rather than central measurements. Appendix B reports profile-likelihood minima of mu0 = -0.036 eV for DESI+PR4 and mu0 = -0.038 eV for DESI+ACT-lite in LambdaCDM, i.e., in the unphysical negative region, so the Bayesian 95% upper limits quoted in the abstract (0.061-0.077 eV) are intersections of a downward-sloping profile with the prior boundary at sum m_nu = 0. The paper states this caveat in Section V and Appendix B, but the abstract and the Section VI discussion do not carry the same qualification, and the phrase 'tightest constraints on the total neutrino mass to date' overstates the robustness of a prior-boundary limit. The abstract and conclusions should be reworded, and the profile-likelihood values in Table IV should be referenced alongside the Bayesian limits so that readers can see the prior dependence directly.","section":"Section V and Appendix B"}],"minor_comments":[{"comment":"The abstract says the ACT-only tension with DESI 'exceeds 3 sigma,' but Section III reports 2.7 sigma with CMB lensing and 3.2 sigma without it; please clarify that the 'exceeding 3 sigma' statement refers to the no-lensing ACT-only configuration.","section":"Abstract and Section III"},{"comment":"The 'Significance' column would benefit from a footnote stating that the n-sigma values are obtained from Delta chi^2 assuming two additional degrees of freedom for w0waCDM relative to LambdaCDM, or from a direct citation of the definition in Ref. [29], so that readers can reproduce the conversion.","section":"Table II"},{"comment":"The statement that 'the errors on w0 and wa remain unchanged compared to P-ACT' should specify that this refers to the 68% uncertainties in Table III, because the contours in Figure 2 have different shapes for DESI+ACT.","section":"Section IV"},{"comment":"Appendix B uses ACT-lite instead of the full MFLike likelihood; this is disclosed in a footnote, but the main text of Section V should mention that the profile-likelihood check is approximate so that readers do not mistake it for a full-likelihood result.","section":"Appendix B"},{"comment":"The abstract's phrase 'increasing to over 4 sigma with the inclusion of Type Ia supernova data' is only literally true for DESY5; Pantheon+ gives 2.8 sigma and Union3 gives 3.7 sigma for the baseline PR4+ACT combination, so the wording should be 'up to over 4 sigma' or 'for the DESY5 sample.'","section":"Abstract and Table II"},{"comment":"The phrase 'contain more information from ACT' is slightly misleading because those combinations also differ in the Planck multipole range; rephrase to 'use ACT over a wider multipole range' for precision.","section":"Figure 6 caption"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this paper does exactly what it says it does. It combines DESI DR2 BAO with the new ACT DR6 CMB data, splicing Planck PR4 and ACT at multipole cuts, and asks whether the DESI dark-energy conclusions hold up. The central result—that the w0waCDM preference over a cosmological constant persists at about 3 sigma, and higher with some supernova samples—looks genuine. The tightest neutrino bound (<0.061 eV) is a new number, and the paper is honest that the bound shifts with the CMB combination and the low-ell EE likelihood.\n\nWhat I credit: the paper is transparent about dataset variations, reports all combinations in tables, and includes a profile-likelihood appendix that makes the prior dependence of the neutrino limit explicit. That is good practice. The heavy reliance on the same collaboration's [16] for definitions and priors is fine; this is a follow-up analysis, and the ACT cross-check is the opposite of circular.\n\nWhere it is softer: the Planck+ACT combination is a cut-based splice with no inter-survey covariance and no relative calibration term. The paper states this clearly, but that approximation is load-bearing. The parameters driving the w0waCDM significance (Omega_ch2, ns) shift with the cut choice, and the neutrino bound moves from <0.061 to <0.077 eV depending on the combination. At a baseline significance of about 3 sigma, an unmodeled calibration offset at the 0.5-sigma level in Omega_ch2 could move the headline below the conventional threshold. That does not sink the paper, but it means the tightest neutrino limit should not be quoted as settled until a joint covariance treatment exists. Minor framing issue: the abstract's 'over 4 sigma' comes only from the DESY5 supernova sample; with Pantheon+ the significance is 2.8 sigma, which the text does report. Also, the data and code are promised but not yet released, so the results are not currently independently reproducible.\n\nBottom line: the paper is a solid, important incremental contribution for anyone tracking the DESI dark-energy question or CMB-neutrino bounds. It deserves a serious referee. The central robustness claim survives the caveats I listed, and the soft spots are the kind a referee should push on, not reasons to reject.","headline":"A careful, transparent robustness check that confirms DESI DR2's evolving-dark-energy preference survives ACT DR6 at about 3 sigma, with the tightest neutrino limit still conditional on the cut-based Planck+ACT splice.","tokens_in":27639,"tokens_out":1845,"would_cite":true,"duration_ms":19806,"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":"This paper argues that DESI's evidence for evolving dark energy—about 3 sigma with CMB data and above 4 sigma when supernovae are added—survives the inclusion of the ACT DR6 measurements, and that a Planck+ACT combination gives the…","keywords":["baryon acoustic oscillations","dark energy","cosmic microwave background","w0waCDM","neutrino mass","cosmological parameter estimation","ACT DR6","Planck PR4"],"falsifier":"Recompute the baseline fit with a modeled Planck–ACT cross-survey covariance, or vary ACT's polarization-efficiency calibration within its quoted uncertainty, and check whether the inferred cold-dark-matter density $\\Omega_{\\mathrm{cdm}} h^2$ shifts enough to move the $w_0w_a$CDM significance below $2\\sigma$; a direct version is an independent higher-precision measurement of $\\Omega_{\\mathrm{cdm}} h^2$ that lands closer to ACT's high value than to the combined PR4+ACT value.","tokens_in":25932,"feed_emoji":"🌌","tokens_out":12413,"duration_ms":114631,"temperature":0.7,"pith_summary":"The paper sets out to test whether the DESI galaxy survey's baryon acoustic oscillation data, which mildly prefer a dark energy that changes over time rather than a constant cosmological constant, hold up when the newest CMB measurements from the ACT telescope are added to Planck. The answer it argues for is yes: the baseline combination of DESI, Planck's PR4 data, and ACT gives about a $3\\sigma$ preference for a time-varying dark energy equation of state, rising above $4\\sigma$ when supernova samples are included. The same combination also produces the tightest neutrino-mass upper limit so far from BAO plus CMB alone, $\\sum m_\\nu < 0.061$ eV at 95% confidence under the standard $\\Lambda$CDM model. The paper matters because the dark-energy question is whether cosmic acceleration is truly constant, and because ACT alone disagrees with DESI at more than $3\\sigma$ in the standard model, so the conflict had to be resolved one way or another.","feed_headline":"Dark energy hint survives ACT telescope's final data","feed_subtitle":"DESI's expansion-rate hint stays near 3 sigma once Planck and ACT are combined; neutrino bound tightens to 0.061 eV.","key_machinery":"The argument runs on the two-parameter dark-energy equation of state $w(a) = w_0 + w_a(1-a)$, with the cosmological constant represented by $(w_0, w_a) = (-1, 0)$, combined with a multipole-cut recipe that joins Planck and ACT power spectra without modeling an inter-survey covariance: Planck supplies $\\ell < 2000$ temperature and $\\ell < 1000$ polarization, ACT supplies the higher multipoles up to $\\ell = 8500$. The cut recipe matters because the value of the cold-dark-matter density $\\Omega_{\\mathrm{cdm}} h^2$ that the CMB combination returns controls how far the dark-energy contours are pulled from $\\Lambda$CDM; ACT alone measures $\\Omega_{\\mathrm{cdm}} h^2$ higher than Planck and is over $3\\sigma$ inconsistent with DESI, while the combined datasets land at a lower $\\Omega_{\\mathrm{cdm}} h^2$ that is only about $2\\sigma$ from DESI, which is what keeps the evolving-dark-energy preference alive.","core_discovery":"In the $w_0w_a$CDM parameterization, $w(a) = w_0 + w_a(1-a)$, the baseline CMB combination—Planck PR4 for multipoles below $\\ell = 2000$ in temperature and $\\ell = 1000$ in polarization, ACT DR6 above those cuts—shifts DESI's dark-energy constraint to a $3.0\\sigma$ preference for $(w_0, w_a)$ away from $(-1, 0)$, and to $2.8\\sigma$, $3.7\\sigma$, and $4.2\\sigma$ when Pantheon+, Union3, and DESY5 supernova samples are added. Across all tested Planck+ACT cut schemes the dark-energy significance changes by at most $0.3\\sigma$, which is the basis for the authors' conclusion that the DESI DR2 results are robust in light of the new ACT data. The neutrino-mass bound is more fragile: under $\\Lambda$CDM with a physical $\\sum m_\\nu > 0$ prior it ranges from $\\sum m_\\nu < 0.061$ eV for the baseline combination to $0.077$ eV for the ACT-preferred P-ACT combination, and the low-multipole polarization likelihood choice can relax limits by up to 7%. These are reported as the tightest BAO+CMB-only neutrino-mass constraints to date.","pith_inferences":["Going beyond the paper: a full covariance between Planck and ACT, including their shared sky fraction and correlated calibration noise, could move $\\Omega_{\\mathrm{cdm}} h^2$ by more than the cut-choice spread the paper sees, and that could in turn shift the $3\\sigma$ dark-energy significance beyond the $0.3\\sigma$ variation reported here.","The cut locations are not determined by a formal optimization, so a fair test of the robustness claim would be to scan the temperature and polarization cut positions and map how the $w_0w_a$CDM significance responds; the paper tests two cut schemes, but the surface between them is unexplored.","Because the dark-energy preference tracks $\\Omega_{\\mathrm{cdm}} h^2$, a higher-precision independent measurement of that density—from small-scale CMB data or from large-scale structure—would either corroborate the evolving-dark-energy interpretation or locate the systematic in the CMB combination."],"forward_implications":["If the central claim is right, the DESI DR2 dark-energy hint is not an artifact of outdated CMB data: adding ACT to Planck keeps the preference for varying dark energy near $3\\sigma$, and above $4\\sigma$ when Union3 or DESY5 supernovae are included.","The multipole-cut Planck+ACT combination sharpens CMB parameter constraints by roughly 14% on the scalar spectral index and 23% on the baryon density compared with Planck alone, so future BAO+CMB analyses can expect modest gains from this combination style.","The tightest BAO+CMB-only neutrino-mass bound becomes $\\sum m_\\nu < 0.061$ eV at 95% confidence under $\\Lambda$CDM, about 5% tighter than the DESI DR2 baseline.","The neutrino-mass result is the fragile part: the 95% upper limit ranges from 0.061 to 0.077 eV depending on which Planck+ACT combination and low-$\\ell$ polarization likelihood is used.","The positive-neutrino-mass preference seen in DESI+CMB data disappears when supernovae are added, in both the Bayesian posterior and the profile-likelihood analysis."],"supporting_citations":[{"why":"Supplies the DESI DR2 BAO measurements and the baseline w0waCDM and neutrino-mass results whose robustness this paper tests.","marker":"[16]"},{"why":"Supplies the ACT DR6 power spectra and the P-ACT combination definition that this paper extends to PR4.","marker":"[31]"},{"why":"Provides the earlier ACT neutrino-mass constraint that this paper updates with DESI DR2 BAO data.","marker":"[33]"},{"why":"Provides the Planck PR4/NPIPE likelihood that anchors the baseline PR4+ACT combination.","marker":"[46]"},{"why":"Provides the Planck low-$ℓ$ Commander and SimAll likelihoods and the PR3 high-$ℓ$ Plik likelihood used in the P-ACT and low-$ℓ$ variants.","marker":"[37, 38]"},{"why":"Supplies the SRoll2 low-$ℓ$ EE likelihood whose choice loosens neutrino-mass bounds by up to 7%.","marker":"[39]"},{"why":"Supplies the ACT DR6 CMB lensing measurement whose inclusion reduces the ACT–DESI tension from 3.2$σ$ to 2.7$σ$.","marker":"[43]"},{"why":"Supplies the DESY5 supernova sample that pushes the evolving-dark-energy significance above 4$σ$.","marker":"[36]"}],"fun_headline_variants":["ACT's final data fails to kill dark energy hint","Planck+ACT keeps DESI dark energy edge at 3 sigma","Neutrino mass bound tightens to 0.061 eV from DESI+CMB","DESI dark energy clue survives ACT+Planck combo","Tighter neutrino bound from combined DESI and CMB data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's central results depend on the assumption that Planck and ACT can be combined by simply cutting each data set to non-overlapping angular scales and treating the two measurements as independent, with no shared calibration error or correlated noise between them.","fun_headline_variants_meta":{"raw":{"variants":["ACT's final data fails to kill dark energy hint","Planck+ACT keeps DESI dark energy edge at 3 sigma","Neutrino mass bound tightens to 0.061 eV from DESI+CMB","DESI dark energy clue survives ACT+Planck combo","Tighter neutrino bound from combined DESI and CMB data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000861,"raw_usage":{"total_tokens":3803,"prompt_tokens":1078,"completion_tokens":2725,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":694,"completion_tokens_details":{"reasoning_tokens":2633}},"tokens_in":694,"tokens_out":2725,"duration_ms":17257,"temperature":1.0,"reasoning_tokens":2633,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:15:56.369217+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the baseline fit with a modeled Planck–ACT cross-survey covariance, or vary ACT's polarization-efficiency calibration within its quoted uncertainty, and check whether the inferred cold-dark-matter density $\\Omega_{\\mathrm{cdm}} h^2$ shifts enough to move the $w_0w_a$CDM significance below $2\\sigma$; a direct version is an independent higher-precision measurement of $\\Omega_{\\mathrm{cdm}} h^2$ that lands closer to ACT's high value than to the combined PR4+ACT value.","supporting_citations":[{"cited_title":"stabilized","cited_arxiv_id":null,"evidence_quote":"Supplies the DESI DR2 BAO measurements and the baseline w0waCDM and neutrino-mass results whose robustness this paper tests."}],"review_version":1}