{"id":"fb7e6f45-b8e2-4491-8c0e-1f3e2cd73383","arxiv_id":"2506.22997","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A two-mediator dark matter model with a near-threshold heavy resonance reconciles relic abundance with self-interactions, but a unit conversion error inflates its direct detection signal by a factor of 1000.","lead":"This paper proposes a dark matter model with two mediator particles: a light one responsible for halo self-interactions, and a heavy one tuned to twice the dark matter mass to boost annihilation in the early universe. The authors claim this resolves the conflict between the observed relic abundance and self-interaction requirements, and they make testable predictions for the LHC and direct detection.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Benchmark violates the model's own mixing relations: y_chi/g_DM = 1.58 forces theta ~ 58 degrees, incompatible with the assumed |theta| << 1 and with a positive-definite scalar mass matrix for m_phi = 15 MeV, m_Phi = 1201 GeV.","rationale":"The strongest-claim benchmark is not merely tuned; it is not a point of the EFT as defined. The paper's central proof of concept is a Table 1 benchmark that satisfies relic and self-interaction constraints, but Eqs. (9)/(A.9) fix y_chi/g_DM = tan(theta). The benchmark values 0.30/0.190 give theta ~ 58 degrees, contradicting the small-mixing assumption and making the light/heavy scalar mass matrix indefinite for the quoted masses. This is an internal inconsistency, not a disagreement with external consensus, and it directly invalidates the claimed demonstration unless the model is reparameterized with independent couplings. The independent-coupling rescue would abandon the natural PNGB origin emphasized throughout the paper, so the central claim as written should be rejected pending redefinition and rescanning. I credit the paper for a detailed numerical implementation and for attempting explicit cosmological checks, and the mechanism of a heavy resonance decoupling freeze-out from late-time self-interactions is plausible; but the concrete benchmark on which the paper rests does not satisfy the model's own field-theory constraints. The reader's weakest assumption (quark-silent phi) is real but secondary; the mixing-relation violation is more load-bearing because it removes the benchmark from the model before external constraints are applied. I also note the Appendix C unit conversion is off by a factor of 1000 (1 GeV^-2 = 3.89e-28 cm^2, not 3.89e-25 cm^2), so the true spin-independent prediction is ~6.7e-51 cm^2, far below next-generation sensitivity; this further weakens the advertised testability but is not the main reason for rejection.","tokens_in":30196,"tokens_out":16779,"duration_ms":189715,"concrete_test":"Recompute the benchmark using the model's own relations: set tan(theta)=y_chi/g_DM from Table 1, insert theta and the quoted masses into Eqs. (A.34)-(A.37), and verify positivity of the two eigenvalues. With theta=57.7 degrees the determinant is negative, so the benchmark is not realizable. If independent couplings are intended instead, the authors must drop Eq. (9) and rerun the Appendix D scan with theta <= 0.1, reporting whether any island with m_phi ~ 15 MeV, sigma_T/m_chi ~ 0.1-1 cm^2/g, and Omega h^2 ~ 0.12 survives.","verdict_should_be":"REJECT","load_bearing_attack":"Eq. (9) and its appendix form (A.9) derive both dark-matter couplings from a single Yukawa y_f after a small mixing rotation: y_chi=(y_f/sqrt(2)) sin(theta) and g_DM=(y_f/sqrt(2)) cos(theta). The Table 1 benchmark has y_chi=0.30 and g_DM=0.190, forcing tan(theta)=1.58, i.e. theta ~ 58 degrees, in direct contradiction to the assumed |theta| << 1 that underlies the PNGB interpretation of the light mediator and the technical smallness of m_phi. Inserting this theta into the scalar mass matrix (A.34)-(A.37) makes the configuration impossible: with m_phi=15 MeV and m_Phi=1201 GeV, the required off-diagonal entry satisfies |m_sa^2| ~ 0.5 m_s^2, so the determinant m_a^2 m_s^2 - (m_sa^2)^2 is negative and the spectrum is not positive-definite. The numerical scan in Appendix D treats y_chi and g_DM as independent parameters and never enforces Eq. (9); the advertised island of viability may therefore live outside the model defined in Section 2.2. This is a more direct threat to the central claim than the unit-conversion and LHC-rate issues noted in the reader's report.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a two-mediator dark-sector EFT in which a light scalar φ (MeV scale) mediates velocity-dependent self-interactions while a heavy scalar Φ_h (TeV scale) near 2m_χ resonantly enhances annihilation during thermal freeze-out, decoupling early-universe relic production from late-time halo dynamics. A numerical scan is claimed to identify a narrow viable island, with a benchmark m_χ = 600 GeV, m_φ = 15 MeV, m_Φh = 1201 GeV, y_χ = 0.30, g_Y1^DM = 0.190 that reproduces Ωh² ≈ 0.120 and σ_T/m_χ ~ 0.1–1 cm²/g at dwarf velocities. The paper also presents LHC, direct-detection, BBN, and indirect-detection signatures, and an optional composite SU(3)_H completion.","tokens_in":30635,"tokens_out":9609,"duration_ms":133259,"significance":"The conceptual mechanism—separating the mediator responsible for SIDM from the mediator responsible for freeze-out—is well motivated and could address a genuine tension in minimal SIDM models. The paper is unusually explicit about its model assumptions and provides extensive appendices, including a radiative-stability check in Appendix G and a detailed scan description in Appendix D. However, the central benchmark is not a point of the model as defined by the paper's own coupling relations in Section 2.2, and the direct-detection prediction contains a factor-1000 unit conversion error. Because the advertised proof-of-concept and one of the headline observables are invalidated, the significance of the paper as it stands is conditional at best.","major_comments":[{"comment":"The Table 1 benchmark is inconsistent with the model's own coupling relations. Equations (9) and (A.9) give y_χ = (y_f/√2) sinθ and g_Y1^DM = (y_f/√2) cosθ, so tanθ = y_χ/g_Y1^DM = 0.30/0.190 = 1.58, i.e. θ ≈ 58°, in direct contradiction to the assumed |θ| ≪ 1 in Eq. (5) and (A.8). Such a mixing angle is also impossible in the mass matrix (A.34)–(A.37) with m_φ = 15 MeV and m_Φh = 1201 GeV: determinant positivity requires |θ| ≲ m_φ/m_Φh ≈ 1.2×10⁻⁵. The numerical scan in Appendix D.2 treats y_χ and g_Y1^DM as independent parameters and never enforces Eq. (9), so the advertised island of viability may lie entirely outside the model defined in Section 2.2.","section":"§2.2, Eqs. (7)–(9); Appendix A.3; Table 1"},{"comment":"The unit conversion in Eq. (C.8) is wrong by a factor of 1000: 1 GeV⁻² = 0.3894×10⁻²⁷ cm² (0.3894 mb), not 0.3894×10⁻²⁴ cm². Therefore the predicted spin-independent cross-section is σ_SI ≈ 6.7×10⁻⁵¹ cm², not 6.7×10⁻⁴⁸ cm². This removes the claimed direct-detection signal from the reach of next-generation experiments and invalidates the corresponding statements in the abstract, Section 5.4, and Section 7.","section":"Appendix C.4, Eq. (C.8)"},{"comment":"The collider production rate is quoted without any calculation. Section 5.5 states an 'O(1–10) fb' rate for σ(pp→Φ_h→ttbar) at √s = 14 TeV, but no gluon-fusion cross-section formula, PDF choice, acceptance estimate, or background analysis is given; Eq. (33) is merely schematic. Since 'within HL-LHC reach' is one of the paper's headline falsifiable predictions, this rate must be substantiated with an actual calculation or a quantitative reference.","section":"§5.5"}],"minor_comments":[{"comment":"There are numerous typographical and grammatical errors, including 'identifed' (Section 1), 'Altough' (Section 4.2), 'is are be specified' (Section 2.2), and inconsistent rendering of 'Schrödinger' in Appendix F; a careful proofread is needed.","section":"Throughout"},{"comment":"The claim that φ 'never thermalizes' is based on Γ/H ~ 0.05–0.15 at T ≈ m_φ, which is not far below unity; the authors should quantify the actual freeze-in abundance and justify the non-thermalization statement with the full Boltzmann yield rather than a rough ratio.","section":"§5.3, Appendix B"},{"comment":"The terminology 'quark-only portal' for Φ_h and 'quark-silent' for φ should be clearly separated at first use to avoid an apparent contradiction: the heavy scalar couples to quarks while the light scalar is leptophilic and quark-silent at tree level.","section":"§5.5, Appendix C.4"},{"comment":"The two-stage scan explicitly tunes the heavy-sector parameters to reproduce Ωh² and selects y_χ and m_φ along the SIDM band; this is a legitimate parameter search, but the language of 'predictive island' should be softened to reflect that the observables are used as inputs to locate the benchmark.","section":"§4.1, Appendix D.2"}],"recommendation":"reject","confidential_remarks":"The determinant-positivity failure of the benchmark under Eq. (9) is decisive: the advertised benchmark cannot be a point of the model as constructed, and the direct-detection conversion error removes another headline signature. Reconciling the model with its own mixing structure would require introducing an independent light-mediator coupling not tied to the PNGB mixing, which is effectively a different model. I would not encourage a resubmission along the current lines."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Honest take: the benchmark that the whole paper rests on is not a point of the model defined in Section 2. Eq. (9) forces tanθ = yχ/g_DM = 0.30/0.190 = 1.58, so θ≈58°. That contradicts the |θ|≪1 used in Eq. (5) and throughout Appendix A.3. If you plug θ≈58° into the mass matrix, the light mass m_φ is still possible, but only by a delicate cancellation between m_a^2 and (m_sa^2)^2/m_s^2; the PNGB naturalness story collapses. The stress-test note claims the determinant goes negative; that specific arithmetic doesn't check out, but the conclusion is the same: the benchmark cannot be realized in the small-mixing EFT. The scan in Appendix D treats yχ and g_DM as independent parameters, so the island of viability probably lives outside the model.\n\nSecond soft spot is the unit conversion in Appendix C.4. 1 GeV^-2 = 0.3894e-27 cm^2, not 0.3894e-24. Corrected σ_SI ≈ 7e-51 cm^2, not 7e-48. So the headline direct-detection prediction is a factor 1000 too optimistic; at 7e-51 cm^2 it is unobservable for the foreseeable future. The LHC rate of O(1-10) fb is stated without any calculation, so we can't verify it. Finally, the 'narrow predictive island' is obtained by tuning m_Phi_h and g_DM to reproduce Ωh²=0.12; that's a standard scan, but calling it 'predictive' overstates what happened.\n\nWhat's genuinely good: the EFT layout is clear, the appendices do real work (non-perturbative Yukawa scattering, BBN constraints for the light scalar), and the optional SU(3)_H UV completion is cleanly separated. The idea of using a single complex scalar to generate both a light PNGB and a heavy radial mode is neat. If the authors can find a benchmark that satisfies Eq. (9) with |θ|≪1, or revise the model to allow a large mixing angle while keeping m_φ technically natural, there could be something here. As it stands, the advertised benchmark doesn't belong to the model, and the direct-detection claim is off by three orders of magnitude.\n\nWho should read it: model builders interested in SIDM with resonant freeze-out might take ideas from the EFT, but they shouldn't rely on the benchmark. Worth a serious referee? I'd say yes, because the flaws are specific and the construction is substantial; a referee can push for the consistency check. But it needs major revision before publication.","headline":"The benchmark that carries the paper violates the model's own mixing relations (θ≈58° vs |θ|≪1), and the direct-detection prediction is off by a factor 1000 due to a unit conversion error; the EFT construction is solid but the advertised island of viability sits outside the defined model.","tokens_in":31181,"tokens_out":19385,"would_cite":false,"duration_ms":164617,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Two dark-sector mediators, one light for halo self-interactions and one heavy resonance at twice the dark-matter mass for freeze-out, satisfy both the relic abundance and galactic constraints, with a predicted 1.2 TeV signal.","keywords":["self-interacting dark matter","resonant annihilation","thermal freeze-out","Sommerfeld enhancement","two-mediator dark sector","composite dark matter","direct detection","density-responsive dark energy"],"falsifier":"A dedicated search for a narrow resonance in the $t\\bar t$ invariant-mass spectrum near 1.2 TeV with the full High-Luminosity LHC dataset: if no peak appears at the predicted production rate of order 1–10 fb, the benchmark's resonant freeze-out mechanism is ruled out. As an independent check, a next-generation direct-detection experiment reaching $\\sigma_{\\rm SI}\\sim10^{-48}$ cm$^2$ that sees no events near $m_\\chi\\simeq600$ GeV would falsify the portal structure the benchmark requires.","tokens_in":29967,"feed_emoji":"🌌","tokens_out":17718,"duration_ms":161220,"temperature":0.7,"pith_summary":"Minimal self-interacting dark matter models face a quantitative conflict: the coupling that produces the observed relic abundance through thermal freeze-out does not produce the self-scattering strength that dwarf galaxies require, and vice versa. The paper argues that a second mediator dissolves this conflict by giving the two jobs to two different particles. A light scalar $\\phi$ with a mass near 15 MeV provides the velocity-dependent attractive Yukawa self-interactions that form cored halos, while a heavy scalar $\\Phi_h$ tuned to the threshold $m_{\\Phi_h}\\approx 2m_\\chi$ boosts the annihilation rate during freeze-out by a factor of about 143 through an $s$-channel resonance. A numerical scan finds a narrow island of viability, with a representative benchmark at $m_\\chi=600$ GeV, $m_\\phi=15$ MeV, and $m_{\\Phi_h}=1201$ GeV that matches the measured relic density and gives $\\sigma_T/m_\\chi$ around 1 to 0.1 cm$^2$/g at dwarf-galaxy velocities while falling four orders of magnitude at cluster scales. If the model is right, it predicts a narrow top-pair resonance near 1.2 TeV within reach of the High-Luminosity LHC and a spin-independent direct-detection cross-section near $7\\times10^{-48}$ cm$^2$.","feed_headline":"One resonance reconciles dark matter's relic and halo constraints","feed_subtitle":"The same model predicts a 1.2 TeV top-pair resonance and a direct-detection signal within next-generation reach.","key_machinery":"The machinery is a scalar pair emerging from one complex field whose potential is spontaneously broken: a heavy radial mode $\\Phi_h$ with mass $\\sqrt{2\\lambda}\\,v_s$ at the TeV scale, and a light pseudo-Nambu-Goldstone boson $\\phi$ at the MeV scale whose small mass is protected by an approximate shift symmetry, mixed by a tiny CP-violating angle $\\theta$. Because the dark fermion $\\chi$ receives its mass from the same vacuum expectation value, the resonance condition $m_{\\Phi_h}\\approx2m_\\chi$ reduces to the coupling relation $\\lambda\\approx y_f^2$, and the small detuning $\\delta=m_{\\Phi_h}/(2m_\\chi)-1\\simeq8.3\\times10^{-4}$ is shown to be radiatively stable. Annihilation proceeds through the $s$-channel Breit-Wigner pole with a total enhancement factor $S_{\\rm total}\\simeq143$ that includes Sommerfeld enhancement from the light mediator; self-interactions are computed non-perturbatively by solving the Schrödinger equation for the attractive Yukawa potential in partial waves.","core_discovery":"The central claim is that the long-standing tension between the thermal relic abundance and the self-interaction requirement is not intrinsic to self-interacting dark matter but a symptom of giving one mediator both jobs. The paper first quantifies the failure of the minimal one-mediator model—at $m_\\chi=100$ GeV and $m_\\phi=20$ MeV, the coupling that gives $\\sigma_T/m_\\chi=1$ cm$^2$/g at 30 km/s overproduces dark matter by a factor of 2.5—and then shows that adding a heavy scalar $\\Phi_h$ near the $2m_\\chi$ threshold restores consistency without changing late-time self-interactions. The proof of concept is a non-empty intersection of the relic-density band and the self-interaction band in the $(m_\\phi, y_\\chi)$ plane, a narrow island spanning roughly $m_\\phi\\in[12,18]$ MeV and $y_\\chi\\in[0.28,0.32]$ at $m_\\chi=600$ GeV once the heavy-sector parameters are fixed by the relic abundance alone.","pith_inferences":["The paper leaves the cosmology of its density-responsive dark-energy sector uncomputed; a dedicated Boltzmann analysis of the $\\rho_\\Phi(X)$ evolution would show whether the mild $w(a)$ deviation it permits also shifts the predicted value of $S_8$, a connection the paper raises but does not quantify.","The quark-silent, leptophilic structure of $\\phi$ occupies a narrow experimental corridor: with $m_\\phi=15$ MeV and $c_e\\simeq5\\times10^{-11}$ the mediator lives about 0.44 s, so tightened nucleosynthesis bounds or new light-scalar beam-dump searches would test this imposed input rather than the resonance mechanism itself.","If the composite SU(3)$_H$ completion is taken seriously, the confinement transition at $\\Lambda_H\\simeq2.5$ TeV should radiate a stochastic gravitational-wave background; estimating its amplitude would convert the optional UV story into a directly testable signature.","The same decoupling logic—a threshold resonance setting the abundance while a light state controls halo physics—suggests that the one-mediator tension is a general structural feature of self-interacting dark matter, not a peculiarity of the Yukawa setup studied here."],"forward_implications":["A narrow scalar resonance at about 1.2 TeV with total width $\\Gamma_{\\Phi_h}\\simeq0.17$ GeV and branching ratio to $t\\bar t$ of 99.85% should appear as a peak in the top-pair invariant mass spectrum, with a production rate of order 1–10 fb at 14 TeV that puts it within the reach of the 3000 fb$^{-1}$ High-Luminosity LHC dataset.","The spin-independent direct-detection cross-section is pinned near $\\sigma_{\\rm SI}\\simeq6.7\\times10^{-48}$ cm$^2$ for the 600 GeV benchmark, just below the current leading limit and inside the projected reach of next-generation liquid-xenon experiments.","The self-interaction cross-section has a sharp, specific velocity dependence—0.96 cm$^2$/g at 10 km/s, 0.11 cm$^2$/g at 30 km/s, and $9.5\\times10^{-5}$ cm$^2$/g at 1000 km/s—so halo and cluster observations can test the model independently of collider results.","Because the resonance operates only at freeze-out velocities, today's annihilation rate is suppressed by a factor of about 143 relative to the canonical thermal value, placing indirect gamma-ray signals safely below current dwarf-galaxy limits.","The viable region scales predictably across $m_\\chi\\in[200,1000]$ GeV as $m_\\phi\\propto m_\\chi^{0.83}$ and $y_\\chi\\propto m_\\chi^{0.51}$, so a detection at any mass fixes the entire family of allowed benchmarks."],"supporting_citations":[{"why":"Defines the minimal one-mediator SIDM framework and the Born and classical scattering cross-sections whose mutual tension motivates the two-mediator construction.","marker":"[6]"},{"why":"Proves the factorization of resonant annihilation and Sommerfeld enhancement that the freeze-out calculation relies on.","marker":"[8]"},{"why":"Numerical package that computes all relic densities and non-perturbative self-interaction cross-sections in the parameter scan.","marker":"[32]"},{"why":"Cosmic-microwave-background measurement that sets the relic-density target $\\Omega h^2 = 0.120\\pm0.001$ the benchmark is tuned to.","marker":"[1]"},{"why":"Supplies the empirical self-interaction strength $\\sigma_T/m\\sim1$ cm$^2$/g at dwarf velocities used as the astrophysical target.","marker":"[5]"},{"why":"Direct-detection experiment whose current limit the predicted $\\sigma_{\\rm SI}\\simeq7\\times10^{-48}$ cm$^2$ sits just below.","marker":"[42]"},{"why":"Lattice study of SU(3) with ten flavors reporting a large mass anomalous dimension, the basis for the $\\gamma\\simeq0.5$ dark-energy input.","marker":"[49]"},{"why":"Lattice-informed mass scaling for the lightest scalar meson that supports the composite origin of $m_{\\Phi_h}/m_\\chi\\approx2$.","marker":"[38]"}],"fun_headline_variants":["Two-mediator model reconciles dark matter relic and halo constraints","Light scalar for halos, heavy resonance for relic: dark matter done right","Resonant annihilation at twice dark matter mass decouples relic and halo","Separate mediators satisfy dark matter's relic and self-interaction needs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model only works if the light mediator is completely silent to quarks and couples to electrons through an extremely small constant, $c_e\\simeq5\\times10^{-11}$, chosen by hand; that input—not the resonance—is what keeps the model within direct-detection and Big Bang nucleosynthesis bounds.","fun_headline_variants_meta":{"raw":{"variants":["Two-mediator model reconciles dark matter relic and halo constraints","Light scalar for halos, heavy resonance for relic: dark matter done right","Resonant annihilation at twice dark matter mass decouples relic and halo","Separate mediators satisfy dark matter's relic and self-interaction needs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000889,"raw_usage":{"total_tokens":3923,"prompt_tokens":1118,"completion_tokens":2805,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":734,"completion_tokens_details":{"reasoning_tokens":2726}},"tokens_in":734,"tokens_out":2805,"duration_ms":26487,"temperature":1.0,"reasoning_tokens":2726,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:54:51.460703+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A dedicated search for a narrow resonance in the $t\\bar t$ invariant-mass spectrum near 1.2 TeV with the full High-Luminosity LHC dataset: if no peak appears at the predicted production rate of order 1–10 fb, the benchmark's resonant freeze-out mechanism is ruled out. As an independent check, a next-generation direct-detection experiment reaching $\\sigma_{\\rm SI}\\sim10^{-48}$ cm$^2$ that sees no events near $m_\\chi\\simeq600$ GeV would falsify the portal structure the benchmark requires.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the minimal one-mediator SIDM framework and the Born and classical scattering cross-sections whose mutual tension motivates the two-mediator construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves the factorization of resonant annihilation and Sommerfeld enhancement that the freeze-out calculation relies on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Numerical package that computes all relic densities and non-perturbative self-interaction cross-sections in the parameter scan."},{"cited_title":"Astrophys.641A6 [Erratum: Astron.Astrophys","cited_arxiv_id":null,"evidence_quote":"Cosmic-microwave-background measurement that sets the relic-density target $\\Omega h^2 = 0.120\\pm0.001$ the benchmark is tuned to."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the empirical self-interaction strength $\\sigma_T/m\\sim1$ cm$^2$/g at dwarf velocities used as the astrophysical target."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Direct-detection experiment whose current limit the predicted $\\sigma_{\\rm SI}\\simeq7\\times10^{-48}$ cm$^2$ sits just below."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Lattice-informed mass scaling for the lightest scalar meson that supports the composite origin of $m_{\\Phi_h}/m_\\chi\\approx2$."}],"review_version":1}