{"id":"8a518ede-1b53-4564-8d29-0b60d7312569","arxiv_id":"2412.13301","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"TeV-scale Higgs-portal dark matter produces Higgs mass corrections larger than the measured Higgs mass, which excludes most such WIMPs except near half the Higgs mass.","lead":"This paper shows that heavy dark matter which couples to the Higgs boson would shift the Higgs mass by quantum corrections large enough to conflict with the measured 125 GeV value, and uses this to push viable Higgs-portal WIMP masses down to a few TeV. For scientists and investors in dark matter searches, it changes what mass range future detectors need to target: most high-mass Higgs-portal WIMPs would already be ruled out if this naturalness argument is accepted.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The m_h→0 stopping rule is a naturalness prior, not an experimental constraint: reproducing M_h=125.20 GeV for scalar DM above ~5–9 TeV requires a negative tree-level quartic, which the paper forbids; the 'ruled out' claim therefore overstates.","rationale":"The paper's numerical pipeline is standard and its two-loop RGE-improved computation with SARAH, SPheno, and micrOMEGAs is a genuine strength; the LZ 2024 comparison is carefully done. The load-bearing weakness is exactly the one the reader identified: the conformal-limit stopping rule converts a fine-tuning preference into a hard exclusion. Since the observed Higgs mass is an input, loop corrections can always be absorbed into the renormalized tree-level quartic; what the paper calls the maximum allowed DM mass is the point where the required counterterm becomes negative (bosonic DM) or violates perturbativity (fermionic DM). The choice to stop at m_h→0 is not forced by any measurement. The numerical bounds of 9.3 TeV, 5.1 TeV, and 4.4 TeV are therefore naturalness bounds, not experimental exclusions. The paper explicitly acknowledges in Sec. 6 that SUSY or tuned cancellations evade the bound, which confirms the criterion's character. A conditional verdict is appropriate: as a model-independent naturalness constraint the paper is a useful contribution, but as a 'ruled out' statement it overstates. My read does not change the reader's verdict.","tokens_in":22001,"tokens_out":8763,"duration_ms":91414,"concrete_test":"Re-run the real singlet scalar scan with SARAH/SPheno + micrOMEGAs, removing the λ_H∈[0,4π] prior and the m_h→0 stop. For each m_DM and relic-satisfying λ_HS, solve for the λ_H (allowing negative values) that yields the two-loop pole mass M_h=125.20 GeV, subject only to |λ_H|<4π and a vacuum metastability lifetime longer than the age of the universe. Record the largest m_DM for which a solution exists and the corresponding fine-tuning measure Δ=|∂log M_h^2/∂log λ_H|. If m_DM,max>9.3 TeV, the headline exclusion is a tuning criterion; if no solution above 9.3 TeV survives, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central exclusion rests entirely on the stopping rule in Sec. 4 Step 5: stop at the conformal limit m_h→0, where m_h^2 = λ_H v^2 and λ_H is restricted to [0,4π] (Sec. 3, after Eq. (5)). For the real and complex singlet scalar models, the DM loop correction to the Higgs mass is positive and grows with m_DM for the relic-fixing λ_HS. To hold the two-loop pole mass at 125.20 GeV for m_DM ≳ 5–9 TeV, one needs a negative tree-level quartic λ_H, i.e., a negative m_h^2 counterterm. The measured Higgs mass does not forbid this: λ_H is an unmeasured renormalized parameter, and the full effective potential can be bounded below even for λ_H<0 when the DM scalar loop contributes a positive h^4 log h^2 term. The paper's λ_H≥0 floor is a boundedness/naturalness preference, not an experimental constraint. The paper itself concedes in Sec. 6 that SUSY or tuned loop cancellations evade the bound. Thus the abstract's 'ruled out in its entirety' is not a consequence of data; it is a prior on acceptable fine-tuning. Quantifying fine-tuning or allowing λ_H<0 could shift the allowed DM mass well beyond 9.3 TeV, up to the perturbativity limits of 30–40 TeV.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper argues that for Higgs-portal dark matter models, the requirement of thermal relic abundance forces the Higgs-DM coupling to grow with DM mass, and the resulting loop corrections to the Higgs mass become so large above a few TeV that the observed Higgs mass 125.20 GeV cannot be reproduced. Using the SARAH/SPheno two-loop spectrum generator and micrOMEGAs, the authors obtain maximum DM masses of about 9.3 TeV (real singlet scalar), 5.1 TeV (complex singlet scalar), and 4.4 TeV (inert doublet), and they claim that, combined with the 2024 LZ limit, the real and complex singlet scalar models are completely ruled out except for a narrow window near M_h/2. The paper explicitly frames the bound as coming from stopping the tree-level Higgs mass at the conformal limit m_h -> 0 with λ_H restricted to [0,4π].","tokens_in":22382,"tokens_out":3140,"duration_ms":31343,"significance":"The computation is transparent and uses standard, machine-checked tools: SARAH, SPheno, and micrOMEGAs, which is a strength of the paper. If the underlying naturalness prior were accepted, the result would sharply constrain a broad class of popular DM models and would be of considerable phenomenological interest. However, the central claim that the heavy DM mass range is 'ruled out' is not a direct consequence of the measured Higgs mass; it is conditional on the prior that the tree-level quartic λ_H must remain non-negative and that no fine-tuned cancellation is allowed. The paper itself concedes in Sec. 6 that SUSY or tuned loop cancellations evade the bound. The analysis of the scalar models at two loops is solid, but the extension to vector and fermionic DM is only at one-loop level with a non-renormalizable operator, so the 'all types of DM' claim is not backed by the same machinery. The result is best read as a naturalness-based upper limit rather than an empirical exclusion.","major_comments":[{"comment":"The central bound is set by the stopping rule 'We stop when we reach the conformal limit of m_h -> 0' combined with the restriction λ_H ∈ [0,4π] stated after Eq. (5). This is a naturalness prior, not an experimental constraint. For m_DM above the quoted limits, one can keep the two-loop pole mass at 125.20 GeV by choosing a negative tree-level λ_H; the measured Higgs mass does not forbid this because λ_H is an unmeasured renormalized parameter. The tree-level boundedness condition λ_H ≥ 0 is not mandatory once the DM one-loop contribution to the effective potential is included, since a positive h^4 log(h^2) term can stabilize the potential even for λ_H < 0. Thus the abstract's 'ruled out in its entirety' overstates what the data imply; the paper actually establishes an upper bound under an explicit no-fine-tuning assumption. This needs to be stated as the first sentence of the abstract and conclusions, and a quantitative fine-tuning measure should be provided so the reader can judge how much of the excluded region is prior-driven.","section":"Sec. 6, 'Before ending...'"},{"comment":"The paper concedes that 'the obvious exception will be the case of SUSY models' and that fine-tuned cancellations between fermionic and scalar loop contributions can evade the bound. This concession substantially weakens the universality claim made in the abstract that the limit 'is applicable to all types of dark matter i.e. scalar, vector, and fermionic, provided they couple directly with Higgs.' Since the paper's own text admits exceptions that are not exotic, the word 'ruled out' should be replaced by 'constrained under the stated naturalness assumptions,' and the exceptions should be quantified rather than relegated to a closing remark.","section":"Sec. 6, paragraph 2"},{"comment":"The analysis of fermionic DM in Sec. 3.2 is performed only at one loop with a fixed renormalization scale μ = m_t and a non-renormalizable operator, while the scalar DM models are treated at two loops with RGE improvement in Sec. 4. The paper's claim that the same limit applies to all DM types is therefore not supported by the same calculational standard. Either the fermionic and vector cases must be run through the same SARAH/SPheno procedure (or an equivalent higher-order treatment), or the claim of universal applicability should be explicitly downgraded to a qualitative expectation. As written, the reader cannot verify that the one-loop, fixed-scale treatment in Fig. 7 is stable under the two-loop corrections that are central to the scalar analysis.","section":"Sec. 3.2 and Sec. 4"}],"minor_comments":[{"comment":"The vector DM one-loop expression in Eq. (13) contains terms of the form -2m_V^2 + 2v^2 λ_HV with a log term 3(2m_V^2 + v^2 λ_HV) log(m_V^2/μ^2); the sign of the finite part relative to the log term should be checked or a reference given, since for large m_V the log term is positive and the finite part is negative, which may affect the direction of the correction.","section":"Sec. 4, Eq. (13)"},{"comment":"The caption of Fig. 2 says the left panel shows 'W (co-annihilation)- and Z-mediated channels'; the text describes W-mediated processes as annihilation rather than co-annihilation. Please clarify the terminology.","section":"Sec. 2, Fig. 2"},{"comment":"The notation for the Higgs-DM coupling alternates between λ_HS, λ_HDM, λ_Hζ, and λ_345 across figures and equations; a single consistent notation with a table of definitions would improve readability.","section":"Sec. 5.1, Eq. (33)"}],"recommendation":"major_revision","confidential_remarks":"The paper has a solid numerical core for the scalar models, but the headline claim is a naturalness constraint presented as an empirical exclusion. The authors should be encouraged to reframe the result as an upper bound under a specified tuning criterion, and to provide a quantitative fine-tuning measure (for example, the Barbieri-Giudice measure or the loop-to-tree ratio) so that the excluded region is not an artifact of a binary prior. If they are unwilling to soften the 'ruled out' language, I would view the manuscript as overclaiming; with a reframing, it could be a useful contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline claim is the thing to argue about. The paper computes two-loop Higgs mass corrections in three Higgs-portal DM models and shows that, if you keep the tree-level quartic λ_H non-negative and stop when the loop correction eats the whole 125 GeV mass (m_h→0), the relic-consistent DM mass is capped at 9.3, 5.1, and 4.4 TeV. That is a new numerical result, and the LZ-2024 combination leaving only the M_h/2 resonance window in the singlet cases is a clean output. The SARAH/SPheno/micrOMEGAs pipeline is standard, the paper states its numerical choices, and the residual scale dependence is at least acknowledged. No circularity: relic density fixes the Higgs-DM coupling, the loop correction is computed from it, and the result is compared to the measured Higgs mass.\n\nThe soft spot is the logical status of the bound. The stopping rule 'm_h→0' with λ_H∈[0,4π] is a naturalness/tuning prior, not a measurement. Nothing in the Higgs mass data forbids a negative tree-level quartic, and the effective potential can still be bounded below. The paper even concedes in Sec. 6 that SUSY or tuned cancellations evade the bound. So 'ruled out in its entirety' overstates. As a naturalness bound, the result is plausible and worth taking seriously; as a data-driven exclusion, it fails. The fix is straightforward: quantify fine-tuning, relax λ_H≥0, and present the results as upper limits under a stated naturalness criterion.\n\nI would send this to a serious referee. The computations are too careful and the implications for the field's experimental program (whether to build multi-ten-tonne detectors for >9 TeV) are too real to desk-reject. The referee should push for a reframing, not a rejection.\n\nYours,","headline":"Solid two-loop bounds, but the 'ruled out' headline only holds under a naturalness prior, not from data.","tokens_in":22921,"tokens_out":2220,"would_cite":true,"duration_ms":21512,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d","14.80.Bn"],"model":"deepseek-v4-flash","headline":"Heavy Higgs-portal dark matter is ruled out by its own Higgs-mass loop corrections.","keywords":["dark matter","Higgs portal","WIMP","Higgs mass correction","naturalness","direct detection","singlet scalar dark matter","inert doublet model"],"falsifier":"A concrete falsifier would be an explicit relic-density-consistent benchmark of the real singlet scalar model with $m_{\\rm DM} = 10$ TeV in which the two-loop MS-corrected Higgs mass equals 125.20 GeV with $\\lambda_H \\in [0,4\\pi]$ and without needing to push $m_h$ to zero; the paper's method excludes such a point, so its existence would refute the claimed upper bound.","tokens_in":2019,"feed_emoji":"🎯","tokens_out":2613,"duration_ms":85465,"temperature":0.7,"pith_summary":"The paper argues that the heavy-mass region of Higgs-portal dark matter, which direct detection experiments struggle to probe because event rates are tiny, can already be excluded by a different route: the same Higgs-dark-matter coupling that sets the relic abundance generates loop corrections to the Higgs mass that grow with the dark matter mass. For a broad class of scalar, vector, and fermionic dark matter models that couple to the Higgs, the paper computes these corrections and shows that relic-consistent dark matter heavier than roughly 5 to 9 TeV would push the loop-corrected Higgs mass above its measured value of 125.20 ± 0.11 GeV, even after the tree-level Higgs quartic is tuned to its most favorable (conformal) limit. It then shows for three concrete models that combining this Higgs-mass bound with the latest LZ (2024) limit rules out the entire parameter space of real and complex singlet scalar dark matter except a narrow window near $M_h/2$, while the inert doublet model retains a viable region up to about 4.4 TeV thanks to co-annihilation. A sympathetic reader would care because this converts an untestable heavy-WIMP regime into a closed one using existing collider data, and it makes a sharp, checkable prediction about where Higgs-portal dark matter may still hide.","feed_headline":"Loop corrections to the Higgs mass close the heavy dark matter window","feed_subtitle":"Heavy Higgs-portal dark matter would push the Higgs mass past 125 GeV; only a narrow near-resonance window survives.","key_machinery":"The load-bearing object is the DM-induced shift in the Higgs mass, $\\delta m_h^2$, computed from the Coleman-Weinberg effective potential; for a scalar DM it is proportional to $\\lambda_{HS} f(m_S^2) + (v\\lambda_{HS})^2 \\log(m_S^2/\\mu^2)$, so it grows with both the Higgs-DM coupling and the DM mass. The paper's chain is: relic abundance fixes $\\lambda_{HS}$ as a function of $m_{\\rm DM}$; this fixes the size of the loop correction; and the observed Higgs mass, together with the allowed range $\\lambda_H \\in [0,4\\pi]$, sets the maximum DM mass once the tree-level contribution $m_h^2 = \\lambda_H v^2$ is reduced to its conformal limit $m_h \\to 0$. The surviving funnel around $M_h/2$ is the narrow window where annihilation through the Higgs is resonantly enhanced, so the required coupling—and hence the loop correction—stays small.","core_discovery":"On its own terms, this paper establishes that DM-induced radiative corrections to the Higgs mass place an upper bound of a few TeV on any non-supersymmetric dark matter candidate whose annihilation is controlled by a Higgs-portal coupling. The logic is a chain of constraints: the observed relic density fixes the Higgs-DM coupling for each DM mass; that coupling, together with the DM mass, fixes the size of the one-loop (and in the detailed models, two-loop) correction to the Higgs mass; and the requirement that the corrected Higgs mass stay at 125.20 ± 0.11 GeV, with the tree-level quartic $\\lambda_H$ allowed only in $[0, 4\\pi]$ and the conformal limit $m_h \\to 0$ as the last stopping point, sets a maximum DM mass. For the real singlet scalar this maximum is 9.3 TeV (7.4 TeV if the DM self-coupling is maximal); for the complex singlet scalar it is 5.1 TeV; for the inert doublet it is 4.4 TeV. Combined with the 2024 LZ direct detection limit, the real and complex singlet scalar models are left only with a narrow band around $M_h/2$, while the inert doublet also keeps a 0.5–4.4 TeV region through co-annihilation.","pith_inferences":["Beyond the paper: the 'ruled out' claim is a naturalness statement. If one allows a negative tree-level Higgs quartic or quantifies fine-tuning instead of forbidding it, relic-consistent Higgs-portal DM above a few TeV reappears; the exclusion would then become a tuning bound rather than an absolute one.","Beyond the paper: the same loop-correction logic should apply to any new scalar, vector, or fermion with sizable Higgs coupling, so the bound can be extended model-by-model as a quick diagnostic before full relic computations are run.","Beyond the paper: the surviving $M_h/2$ window is a concrete target—a future direct detection experiment with sensitivity at the resonance mass, or a precise Higgs invisible-width measurement, can test whether that window is populated or empty.","Beyond the paper: the authors' focus on non-SUSY models suggests a natural counter-check, namely that supersymmetric spectra with cancellations between fermion and scalar loops would evade the bound, making the mechanism a potential discriminator between SUSY and non-SUSY Higgs-portal dark matter."],"forward_implications":["For the real singlet scalar model, relic-consistent dark matter above 9.3 TeV ($\\lambda_S \\approx 0$) or 7.4 TeV ($\\lambda_S = \\sqrt{4\\pi}$) is excluded by the Higgs mass bound alone.","Adding the LZ (2024) limit closes the real and complex singlet scalar parameter space except for a narrow resonance window near $M_h/2$.","The complex singlet scalar upper bound falls from about 30 TeV (naive perturbativity) to about 5.1 TeV once two-loop Higgs mass corrections are imposed.","The inert doublet model keeps a viable high-mass region, roughly 0.5–4.4 TeV, because W and Z co-annihilation channels relax the required Higgs-DM coupling; above 4.4 TeV even co-annihilation cannot save it.","For any non-SUSY Higgs-portal DM (scalar, vector, or fermion) whose relic density is set by the same coupling, the generic DM mass ceiling is a few TeV—much stricter than the unitarity ceiling of ~100 TeV or perturbativity ceilings of 30–40 TeV."],"supporting_citations":[{"why":"Supplies the Planck relic density range (0.1126–0.1246) that fixes the Higgs-DM coupling as a function of DM mass.","marker":"[4]"},{"why":"Sets the unitarity ceiling (~100 TeV) that the Higgs-mass bound is compared against.","marker":"[8]"},{"why":"Provides the measured Higgs mass 125.20 ± 0.11 GeV and the oblique parameters used as constraints.","marker":"[33]"},{"why":"Gives the more accurate S-matrix unitarity limit on scalar couplings, which the paper contrasts with its conservative naive perturbativity cut.","marker":"[34]"},{"why":"Supplies the 4.2 tonne-year LZ (2024) direct detection limit that closes the parameter space when combined with the Higgs-mass bound.","marker":"[38]"},{"why":"Provides the Coleman-Weinberg effective potential method for computing one-loop corrections to the Higgs mass.","marker":"[43]"},{"why":"SARAH generates the model files used for the two-loop Higgs mass computation.","marker":"[49]"},{"why":"SPheno computes the two-loop MS Higgs mass corrections and renormalization-group evolution.","marker":"[51]"},{"why":"micrOMEGAs computes DM relic density and DM-nucleon scattering cross-sections for the scanned parameter points.","marker":"[53]"}],"fun_headline_variants":["Heavy dark matter hits a Higgs mass ceiling","Higgs mass corrections shut the door on heavy dark matter","Dark matter heavier than 9 TeV breaks the Higgs mass","Higgs-portal dark matter hits a mass wall at 9 TeV","Loop effects turn off heavy Higgs-portal dark matter"],"cache_read_input_tokens":24960,"weakest_assumption_plain":"The exclusion holds only if the tree-level Higgs self-coupling $\\lambda_H$ is required to stay in $[0,4\\pi]$ and the computation is stopped at the conformal limit $m_h \\to 0$; admitting a negative or fine-tuned $\\lambda_H$ would let the same relic-consistent dark matter be heavier than a few TeV.","fun_headline_variants_meta":{"raw":{"variants":["Heavy dark matter hits a Higgs mass ceiling","Higgs mass corrections shut the door on heavy dark matter","Dark matter heavier than 9 TeV breaks the Higgs mass","Higgs-portal dark matter hits a mass wall at 9 TeV","Loop effects turn off heavy Higgs-portal dark matter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000212,"raw_usage":{"total_tokens":1452,"prompt_tokens":1011,"completion_tokens":441,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":627,"completion_tokens_details":{"reasoning_tokens":358}},"tokens_in":627,"tokens_out":441,"duration_ms":4332,"temperature":1.0,"reasoning_tokens":358,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:16:06.769397+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete falsifier would be an explicit relic-density-consistent benchmark of the real singlet scalar model with $m_{\\rm DM} = 10$ TeV in which the two-loop MS-corrected Higgs mass equals 125.20 GeV with $\\lambda_H \\in [0,4\\pi]$ and without needing to push $m_h$ to zero; the paper's method excludes such a point, so its existence would refute the claimed upper bound.","supporting_citations":[{"cited_title":"freeze-out","cited_arxiv_id":null,"evidence_quote":"Supplies the Planck relic density range (0.1126–0.1246) that fixes the Higgs-DM coupling as a function of DM mass."},{"cited_title":"Loop-Level","cited_arxiv_id":null,"evidence_quote":"Sets the unitarity ceiling (~100 TeV) that the Higgs-mass bound is compared against."},{"cited_title":"Testing a lepton quarticity flavor theory of neutrino oscillations with the DUNE experiment","cited_arxiv_id":"1711.10318","evidence_quote":"Provides the Coleman-Weinberg effective potential method for computing one-loop corrections to the Higgs mass."}],"review_version":1}