{"id":"72a95f13-1ffd-4f26-bfce-52795a35f57d","arxiv_id":"2502.03390","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":9,"one_line_summary":"The first Dark Scattering constraints from DES Y3 plus BOSS DR12 give w = -1.04+0.10-0.08 and Ads = -0.2+4.6-5.9 b GeV^-1, consistent with no interaction while reducing S8 projection effects.","lead":"This paper measures how well the Dark Scattering model, in which dark energy and dark matter exchange momentum, fits the latest galaxy survey data from DES and BOSS. It finds that combining the two surveys constrains the model better than either alone and that the model can match both early-universe and late-universe measurements without using CMB data.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central 'solution' claim rests on MAP values for a model the paper's own evidence disfavors; posterior-based tests are needed before claiming DS reconciles CMB and LSS.","rationale":"The paper's headline contribution is the claim that DS can reconcile early- and late-time measurements. The authors' own numbers undermine this: the model is not detected, the marginalized S8 is still low, and the Bayes factor disfavors DS. The only support is a MAP comparison for As, but MAPs in highly non-Gaussian, degenerate posteriors (which they acknowledge via projection effects) are unstable and prior-dependent. The modelling assumption flagged by the reader (EFTofLSS rescaling) is also a concern, but even if that rescaling were exact, the Section 4 claim would still not follow from the reported posteriors. Therefore the load-bearing weakness is the interpretation, not the modelling per se. A joint CMB+LSS fit with a posterior tension metric would settle it. This does not require rejecting the paper; the constraints in Table 1 are useful. But the 'solution' language should be conditional on the posterior test.","tokens_in":10845,"tokens_out":5656,"duration_ms":55026,"concrete_test":"Perform a true joint fit adding the Planck PR4 likelihood (without lensing) to the DES Y3 + BOSS DS analysis, and compute the posterior distribution of S8 and the Bayesian evidence relative to ΛCDM. If the joint S8 posterior still differs from the CMB-only constraint by >1σ, or if ΔlnZ remains negative, the claimed 'solution' in Section 4 is not supported. This directly tests whether DS can consistently connect the two datasets rather than relying on MAP values.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4 claims DS 'offers a solution that can consistently connect' CMB and LSS, based on ln(10^10 As)=3.029 at the DS MAP versus 3.035±0.014 from Planck PR4, while ΛCDM gives 2.975. This is a point-estimate comparison inside a highly non-Gaussian posterior with strong projection effects (the 'butterfly' in w-Ads, Figure 3). The same Table 1 shows the joint DS marginalized S8=0.790±0.018 remains below the CMB value, Ads is consistent with 0 at <1σ, and logZ = 5627.8 vs 5630.5 for ΛCDM, i.e., the evidence disfavors DS by ΔlnZ≈−2.7. Thus the 'solution' is not supported by the posterior or model comparison; it is a MAP artifact. If the MAP shifts under alternative EFT nuisance priors or minimization choices, the central claim disappears.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper presents constraints on the Dark Scattering (DS) interacting dark energy model, characterized by parameters (w, Ads) with Ads=(1+w)ξ, using DES Y3 3x2pt measurements, BOSS DR12 full-shape power spectrum multipoles plus external BAO data, and their joint combination, with a Planck PR4 CMB analysis for comparison. The analysis uses a DS emulator and the halo-model reaction framework for nonlinear clustering, and compares a direct joint likelihood run with a normalizing-flow combination of the individual posteriors. The main quantitative results are that the joint LSS analysis yields w=-1.04+0.10-0.08 and Ads=-0.2+4.6-5.9 b GeV^-1, consistent with ΛCDM, and that the joint combination reduces projection effects relative to single probes. The authors also note that MAP-derived values of the primordial amplitude in DS are closer to the Planck PR4 value than in ΛCDM, which they interpret as evidence that DS can reconcile CMB and LSS measurements. The reported Bayes factor, however, mildly disfavors DS relative to ΛCDM in the joint analysis (ΔlnZ≈-2.7).","tokens_in":11063,"tokens_out":8616,"duration_ms":76238,"significance":"The paper is a competent data-analysis contribution: it provides the first DS constraints from DES Y3, the first joint DES Y3+BOSS DR12 full-shape DS constraints, and a useful stress test of normalizing-flow posterior combination for a highly non-Gaussian extended cosmology. Strengths include the use of public data and pipelines, an independent DS emulator, a halo-model reaction calibrated against DS N-body simulations, and explicit reporting of MAP versus marginalized values to diagnose projection effects. If the central 'solution to the S8 tension' claim were supported, the result would be important; however, the paper's own numbers show no significant detection of DS and negative evidence relative to ΛCDM, so the realistic significance is that DS remains a viable but unrequired extension. With the interpretation tempered, the paper is of interest to the LSS cosmological analysis community, particularly for Stage IV survey preparations.","major_comments":[{"comment":"The sentence 'From this we see that DS offers a solution that can consistently connect early-time measurements of the matter density fluctuations in the CMB with late-time LSS measurements' is not supported by the posterior quantities in Table 1. The joint DS analysis gives Ads=-0.2+4.6-5.9 b GeV^-1, w=-1.04+0.10-0.08, and logZ=5627.8 versus 5630.5 for ΛCDM, i.e., the evidence disfavors DS; the MAP value Ads=0.0 lies at the ΛCDM limit. Comparing ln(10^10 As)=3.029 at the DS MAP to the Planck PR4 value 3.035±0.014 is a point-estimate comparison inside a non-Gaussian posterior with strong projection effects, not a posterior statement. Please replace this conclusion with a posterior-based comparison (e.g., the posterior probability of S8 or As under each model, or a profile likelihood), or explicitly downgrade the claim to a hint that is currently disfavored by the evidence.","section":"Section 4"},{"comment":"The statement that 'it is sufficient to re-scale the standard perturbative components of the power spectrum multipoles by the modified growth factor and rate' is an assumption about the EFTofLSS modelling for DS, and it is load-bearing for the FS+BAO and joint constraints. Because the DS modification is scale-independent at linear level but the halo-model reaction is scale-dependent, it is not obvious that a single rescaling of the one-loop EFT components absorbs all scale-dependent corrections at the k-ranges used. Please provide a validation against N-body simulations or a quantitative estimate of the systematic error this introduces in S8 and Ads; alternatively, state clearly in the conclusions that the constraints are conditional on this rescaling.","section":"Section 3"},{"comment":"For the FS+BAO DS analysis, Table 1 and Figure 2 report two MAP solutions with very similar chi2 values (e.g., Omega_m=0.309|0.311 and S8=0.711|0.864), and the text states 'the data are not constraining enough to draw any conclusion about the DS constraints in this analysis setup.' This bimodality is relevant to the joint analysis: the reported joint constraints and the reduction of projection effects may depend on which FS+BAO mode is connected to the DES posterior. Please quantify the robustness of the joint DS constraints to the choice of mode (e.g., by running the joint MCMC from both MAP solutions and reporting the resulting marginalized intervals).","section":"Section 4 and Table 1"}],"minor_comments":[{"comment":"The phrase 'unconstrained by CMB data' is imprecise: the PR4 DS analysis in this paper shows that w and Ads are unconstrained by CMB data, while other parameters such as omega_b and n_s are constrained. Please rephrase.","section":"Abstract"},{"comment":"The notation '0.309 | 0.311' and the 'second-best MAP' values in Figure 2 are not defined in the text or caption; please add a note explaining that these are the two local maxima found by the minimizer.","section":"Table 1"},{"comment":"The paper says 'we use the same scale-cuts as in the DES Y3 ΛCDM baseline' but uses HMCode2020 for ΛCDM and the DS-emulator for DS; please state whether these scale cuts were validated for the DS-emulator or merely inherited from the ΛCDM analysis.","section":"Section 3"},{"comment":"In the sentence beginning 'The effects of enlarging the prior are small in ΛCDM', the quantitative claim would benefit from a reference to the contours in Figure 3 or a table; as written, the reader cannot assess the size of the shifts.","section":"Section 4"},{"comment":"The definition of the 'boosted posteriors' used for normalizing-flow training is only described in prose; please specify the boost factor and the number of samples used, for reproducibility.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is technically sound in its likelihood setup, but the headline claim is an overinterpretation of MAP values. I recommend major revision: the authors should either provide a posterior-based reconciliation test or substantially soften the 'solution' language. The EFT rescaling assumption should also be validated or clearly caveated. If these are addressed, the paper will be a solid constraints contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is worth a serious read. It delivers the first Dark Scattering constraints from DES Y3 3x2pt, the first joint DES Y3 + BOSS DR12 full-shape analysis, and a PR4 CMB run for DS. The execution is careful: the FS+BAO part reproduces Carrilho et al. (2023), the DES re-analysis matches Abbott et al. (2023a) in Lambda CDM, and the normalizing-flow combination is stress-tested on a genuinely non-Gaussian, projection-dominated posterior. That is real work, and the joint analysis does something useful: it brings MAP and marginalized values closer and weakens sensitivity to EFT nuisance priors.\n\nThe soft spot is the interpretation. The paper says DS 'offers a solution that can consistently connect' CMB and late-time LSS, but that rests on a MAP-value comparison of ln(10^10 As). The marginalized posterior tells a different story: Ads is consistent with zero at <1 sigma, the joint DS S8 = 0.790 +/- 0.018 is still below the CMB value, and the Bayes factor actually disfavors DS by Delta ln Z about -2.7 relative to Lambda CDM. The authors do note no significant detection, but the 'exemplary solution' language overstates what a point estimate in a butterfly-shaped posterior can support. The claim should be rephrased as 'consistent with a possible resolution, subject to projection effects,' or backed by a posterior-based tension metric.\n\nA second, smaller concern: the EFTofLSS rescaling for DS is asserted, not derived. The argument that the linear growth modification is scale-independent is plausible, but a short validation against an N-body-based emulator or a scale-dependent check would have tightened the modeling. This is not fatal, but it is the kind of assumption a referee should ask to see defended.\n\nCitation pattern looks fine; the priors on Ads are motivated by earlier DS work from overlapping authors, but that is a prior choice, not a circularity.\n\nVerdict: conditional acceptance. The paper is a competent, useful constraints analysis with a modest overinterpretation at the end. A revision that demotes the MAP-based claim and adds a few sentences on the rescaling assumption would make it solid. I would bring it to reading group and cite it.\n\nRecommendation: send to peer review.","headline":"New DS constraints are solid; the 'solution' claim is a MAP artifact that needs toning down.","tokens_in":11647,"tokens_out":2596,"would_cite":true,"duration_ms":24625,"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":"Dark Scattering survives joint DES and BOSS full-shape constraints and can connect CMB to late-time structure measurements.","keywords":["dark scattering","interacting dark energy","S8 tension","DES Y3 3x2pt","BOSS full-shape BAO","EFTofLSS","normalising flows","cosmic shear"],"falsifier":"Use Dark Scattering N-body simulations to compute the true galaxy power spectrum multipoles at the BOSS scale cuts without the rescaling approximation; if the best-fit S8 and Ads from those simulated spectra differ from the paper's constraints by more than the reported uncertainties, the central claim collapses.","tokens_in":10673,"feed_emoji":"🔭","tokens_out":6168,"duration_ms":56915,"temperature":0.7,"pith_summary":"This paper tries to establish that Dark Scattering, an interacting dark energy model with pure momentum exchange between dark energy and dark matter, remains a viable explanation of the S8 tension when tested against current galaxy-survey data. Using DES Y3 3x2pt measurements, BOSS DR12 full-shape plus BAO data, and their joint analysis, it finds no significant detection of Dark Scattering but constraints consistent with the ΛCDM limit. The central result is that combining the two large-scale-structure probes reduces projection effects without needing CMB information, bringing marginalized posterior maxima closer to the best-fit values. The authors argue that Dark Scattering offers a solution that consistently connects early-time CMB measurements of matter fluctuations with late-time LSS measurements, with the ΛCDM tension reduced to about 1.3 sigma against Planck PR4. This matters because forthcoming Stage IV surveys will need to test whether momentum exchange can resolve the apparent disagreement.","feed_headline":"Dark scattering keeps S8 consistent from CMB to galaxies","feed_subtitle":"DES Y3 plus BOSS DR12 full-shape data cut projection effects and match CMB-informed results.","key_machinery":"Dark Scattering is the central object: an interacting dark energy model in which dark energy and cold dark matter exchange momentum via elastic scattering, with no energy transfer, leaving the expansion history unchanged. The interaction appears as a drag term in the linearised Euler equation for dark matter, controlled by the combined parameter Ads = (1+w)xi, where w is the dark energy equation of state and xi is the ratio of scattering cross-section to dark matter mass; the S8-relevant regime is Ads > 0 and w > -1, which suppresses linear growth at late times. The analysis machinery combines the halo-model reaction framework for nonlinear structure growth, the EFTofLSS power-spectrum multipoles rescaled by the DS-modified growth factor and growth rate, and normalising flows for combining posterior distributions of independent probes.","core_discovery":"The central claim is that joint large-scale-structure data alone can constrain Dark Scattering nearly as well as a single-probe analysis with CMB-informed priors, and that Dark Scattering remains a viable resolution of the S8 tension. The joint DES Y3 + BOSS DR12 analysis yields DS parameters w = -1.04+0.10-0.08 and Ads = -0.2+4.6-5.9 b $GeV^{-1}$, with no significant detection. The authors state that DS offers a solution that can consistently connect early-time measurements of the matter density fluctuations in the CMB with late-time LSS measurements, because the MAP values of the primordial amplitude in the joint analysis agree with Planck PR4 in DS, while in ΛCDM they do not. The combination of probes also brings marginalized posterior maxima closer to the best-fit values, indicating reduced projection effects, and weakens sensitivity to the priors of the EFTofLSS nuisance parameters.","pith_inferences":["A direct test of the paper's rescaling assumption would be to compare EFTofLSS predictions against Dark Scattering N-body simulations at the exact BOSS scale cuts; any mismatch would shift the inferred Ads and S8.","The scale-dependent enhancement of structure growth inside collapsed halos predicted by Dark Scattering could show up in Stage IV cluster counts or small-scale galaxy-galaxy lensing, providing an independent probe.","The persistent 'butterfly' posterior in the w–Ads plane hints at a fundamental degeneracy that might only be broken by CMB lensing or peculiar-velocity data.","If future data push Ads away from zero with w > -1 at high significance, Dark Scattering would become a preferred resolution to S8; if not, the ΛCDM limit will hold."],"forward_implications":["If joint LSS data can constrain Dark Scattering without CMB information, Stage IV surveys can test the model without relying on CMB assumptions.","The reduced projection effects in the joint analysis make the marginalized constraints more trustworthy guides to the true parameter values.","A non-detection of Dark Scattering with current data does not rule it out as the S8-tension resolution, since the MAP values remain consistent with the CMB.","The normalising-flow combination survived a stress test on highly non-Gaussian, projected posteriors, making it a viable tool for future joint analyses of extended cosmologies."],"supporting_citations":[{"why":"Introduces the Dark Scattering model with pure momentum exchange between dark energy and dark matter.","marker":"Pourtsidou et al. 2013"},{"why":"Showed that Dark Scattering leaves CMB constraints nearly unaffected, motivating it as a CMB-safe S8 resolution.","marker":"Pourtsidou & Tram 2016"},{"why":"Provided the previous BOSS full-shape Dark Scattering constraints and the priors/setup used in the FS+BAO analysis.","marker":"Carrilho et al. 2023"},{"why":"Supplied the DS-emulator used in CosmoSIS for fast computation of linear and nonlinear matter power spectra.","marker":"Carrion et al. 2024"},{"why":"Introduced the halo model reaction framework used to model nonlinear structure growth in Dark Scattering.","marker":"Cataneo et al. 2019"},{"why":"Developed and tested the reaction framework for modified gravity and dark scattering, supporting the nonlinear DS modelling.","marker":"Bose et al. 2020"},{"why":"Provided the BOSS DR12 power spectrum multipoles and covariance matrix used in the full-shape analysis.","marker":"Philcox & Ivanov 2022"},{"why":"Released the DES Y3 3x2pt measurements and baseline analysis choices that the photometric analysis follows.","marker":"Abbott et al. 2022b"},{"why":"Provided the Planck PR4 likelihoods used for the CMB comparison and the quoted S8 tension value.","marker":"Tristram et al. 2024"},{"why":"Developed the EFTofLSS modelling module used in CosmoSIS for the BOSS power spectrum multipoles.","marker":"Moretti et al. 2023"}],"fun_headline_variants":["Joint DES+BOSS data fit dark scattering, ease S8 tension","No CMB needed: dark scattering matches Planck in full-shape data","Full-shape data alone constrain dark scattering and ease S8","Joint DES-BOSS data reduce projection effects, support dark scattering"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The analysis assumes that the effect of Dark Scattering on the mildly nonlinear galaxy power spectrum is fully captured by rescaling the standard perturbation-theory components by the DS-modified linear growth factor and growth rate, plus the halo-model reaction framework; if scale-dependent corrections are missed at the adopted scale cuts, the inferred S8 and Ads values will shift.","fun_headline_variants_meta":{"raw":{"variants":["Joint DES+BOSS data fit dark scattering, ease S8 tension","No CMB needed: dark scattering matches Planck in full-shape data","Full-shape data alone constrain dark scattering and ease S8","Joint DES-BOSS data reduce projection effects, support dark scattering"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001137,"raw_usage":{"total_tokens":4740,"prompt_tokens":985,"completion_tokens":3755,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":601,"completion_tokens_details":{"reasoning_tokens":3681}},"tokens_in":601,"tokens_out":3755,"duration_ms":24725,"temperature":1.0,"reasoning_tokens":3681,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T04:52:46.383615+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Use Dark Scattering N-body simulations to compute the true galaxy power spectrum multipoles at the BOSS scale cuts without the rescaling approximation; if the best-fit S8 and Ads from those simulated spectra differ from the paper's constraints by more than the reported uncertainties, the central claim collapses.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the Dark Scattering model with pure momentum exchange between dark energy and dark matter."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplied the DS-emulator used in CosmoSIS for fast computation of linear and nonlinear matter power spectra."}],"review_version":1}