{"id":"fa320cec-81bf-4ab4-aafd-c9483cf78a13","arxiv_id":"2505.07753","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In two-singlet scalar dark matter models, a light (125-230 GeV) component with tiny relic density can evade direct detection while a heavy (4-10 TeV) partner supplies the observed abundance, but LZ 2024 nearly closes this window.","lead":"This paper maps the allowed mass ranges for dark matter in models that add one, two, or three real scalar singlet fields to the Standard Model. It finds a new window in the two-singlet model where one dark matter particle is light enough to be produced at the LHC, while a heavier partner provides most of the dark matter.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central N-singlet stability claim is unsupported: the restricted three-singlet scan omits all-heavy/all-light hierarchies, and fraction-weighted DD already opens new regions without new annihilation channels.","rationale":"The reader's weakest assumption and my load-bearing concern coincide: the three-singlet conclusion is extrapolated from a restricted scan. The paper's two-singlet new mass window is a separate, tractable scan result and is not invalidated by this concern, so the reader's CONDITIONAL verdict remains appropriate. The all-heavy hierarchy is the sharpest test because the mechanism that created the one-light-two-heavy region is fraction-weighted DD exclusion, which does not require any new annihilation channel; the same mechanism could operate with three heavy states. The manuscript's own concluding sentence that adding more singlets makes bounds increasingly loose reinforces the need for this check rather than replacing it. I considered the strong-coupling regime of the benchmark points (κH1 up to about 10) and the corrupted, duplicated text in the uploaded version as secondary concerns, but they do not change the verdict: the former is partially addressed by the quoted perturbative-unitarity conditions, and the latter affects reproducibility rather than the logical structure of the central claim. The recommended concrete test would settle whether the 'no dramatic change' assertion survives.","tokens_in":26195,"tokens_out":6461,"duration_ms":64872,"concrete_test":"Run the SM+3RSS model with micrOMEGAs 6.1 in the all-heavy hierarchy, e.g. mS1 < mS2 < mS3 all within [4, 10] TeV, scanning κHr over [1, 10] and λij over [10^-4, 1] with the same Planck, LZ 2024 DD, LHC invisible-Higgs, and theoretical constraints as in Sec. 3.1, using the same parameter ranges as the one-light-two-heavy scan. If any allowed points appear with all three relic-density fractions comparable and with masses or couplings outside the ranges reported in Sec. 3.1, the 'no dramatic change' conclusion fails; if none appear, the claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's claim that 'adding more singlets ... will not change this picture dramatically' is the least secure part of the argument. Section 3.1 restricts the three-singlet scan to two hierarchies: two-light-one-heavy with mS3 < mS1 < mS2 and one-light-two-heavy with mS1 < mS2 < mS3, justified by 'no new DM (co-)annihilation channels opening up'. That justification is insufficient. The new one-light-two-heavy allowed region is itself produced without any new depletion channel: it arises because DD exclusion for each heavy state scales with its relic-density fraction, so two heavy states sharing the abundance relax the DD bound (Sec. 3.1, Eq. (3.7) and the following text). By the same mechanism, an unscanned all-heavy hierarchy with three states above roughly 1 TeV sharing the relic density could further relax DD constraints and shift the allowed masses and couplings. The paper's own conclusion states 'Adding more singlets would make the bounds on masses and portal couplings increasingly loose', which is precisely a quantitative change and sits in tension with the 'no dramatic change' message. The manuscript itself admits: 'although our scans were performed for particular regions of the parameter space there are no obvious physical reasons to expect the new allowed regions would arise.' This is a self-acknowledged limitation, not a proof of stability. The central two-singlet window is a defensible scan result, but the N-singlet extrapolation rests on an unverified assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies real scalar singlet extensions of the Standard Model with unbroken Z2 symmetries, using micrOMEGAs for relic-density and direct-detection constraints and MadGraph for LHC mono-X cross sections. It first revisits the one-singlet model, then analyzes two real singlets with independent Z2 symmetries, two real singlets with a single common Z2, and three real singlets with three independent Z2 symmetries. The main claimed results are: (i) in the one-singlet model, DM is excluded below about 3.5 TeV except near the Higgs resonance; (ii) adding a second independently-odd singlet opens a new allowed window with a light state mS1 in [124.8, 230.0] GeV carrying a tiny relic fraction Omega_S1 h^2 ~ 10^-7 and a heavy state mS2 in [4321, 9977] GeV carrying essentially all of the relic density; (iii) a three-singlet configuration with one light and two heavy states further relaxes direct-detection constraints because the two heavy states share the relic density; and (iv) the same-Z2 two-singlet model allows the DM mass to span a much wider range because the DD coupling and the relic-density-controlling couplings are independent after mass-diagonalization. The paper also evaluates LHC mono-jet, mono-Higgs, mono-Z, and di-jet b-bbar signatures for benchmark points.","tokens_in":26574,"tokens_out":9950,"duration_ms":107553,"significance":"If the central two-singlet result holds, it identifies a genuinely interesting phenomenology: a two-component scalar DM model with a light LHC-accessible state that evades direct detection through its small relic-density fraction. The treatment of the direct-detection bound for subdominant components, sigma_SI x Omega_Sr/Omega_DM, is the correct standard scaling, and the paper uses current constraints including LZ 2024. The use of public tools (micrOMEGAs 6.0/6.1, MadGraph5_aMC@NLO) and the explicit benchmarking against ATLAS mono-X analyses are strengths. The significance is tempered, however, by the fact that the paper's most general claims, in particular the abstract's statement that adding more independent-Z2 singlets will not change the picture dramatically, rest on a restricted three-singlet scan and on an under-documented scan of the same-Z2 model. The robust, well-supported part is the two-singlet new mass window; the N-singlet extrapolation needs either additional scans or a softened claim.","major_comments":[{"comment":"The statement that adding more independent-Z2 singlets 'will not change this picture dramatically' is not supported by the scans presented. Section 3.1 restricts the three-singlet scan to two hierarchies, mS3<mS1<mS2 and mS1<mS2<mS3, with the justification that no new (co-)annihilation channels open. That justification is insufficient: the newly found one-light-two-heavy allowed region is itself produced without any new depletion channel. It arises, as the text around Eq. (3.7) states, because the direct-detection exclusion for each heavy state is scaled by the relic-density fraction Omega_Sr/Omega_DM. The same fraction-weighting mechanism implies that an unscanned all-heavy hierarchy, with three states above roughly 1 TeV each carrying Omega ~ 1/3, could further relax the DD tension and shift the allowed masses downward relative to the one-singlet lower bound; an all-light hierarchy with multiple small relic fractions is likewise not excluded by the stated argument. The manuscript itself concedes after Eq. (3.7) that 'there are no obvious physical reasons to expect the new allowed regions would arise,' which is a self-acknowledged limitation, not a proof of stability. To support the abstract's N-singlet claim, the authors should scan the all-heavy and all-light hierarchies, and ideally a four-singlet one-light-three-heavy case, or alternatively restrict the claim to the two scanned hierarchies. The final sentence of Sec. 5, 'Adding more singlets would make the bounds on masses and portal couplings increasingly loose,' is in tension with the abstract's 'not dramatically' phrasing and should be reconciled.","section":"Sec. 3.1 and abstract; Sec. 5 conclusion"},{"comment":"The abstract's claim that adding singlets all odd under the same Z2 allows the DM mass to span the entire range from half the Higgs mass to the TeV scale is not verifiable from the material presented. The text argues that one can make the DD-relevant coupling kappa_H1 small while kappa_H2 and kappa_H12 are large enough to set the relic density, but the effectiveness of kappa_H2 and kappa_H12 depends crucially on the mass splitting between chi2 and chi1: coannihilation requires the states to be close in mass, while a large splitting suppresses the heavier state's abundance and its contribution to depletion. The mass spectrum, the values or ranges of m_chi2, and the scan ranges for the couplings in Eqs. (2.13)-(2.20) are not stated. Without this information, the 'entire mass range' conclusion and the representative plots in Figs. 9-10 cannot be independently assessed.","section":"Sec. 2.2, Eqs. (2.9)-(2.20), Figs. 9-10"},{"comment":"The text states that the two-singlet parameter space was scanned 'exhaustively and uniformly,' but no scan ranges, step sizes, number of points, or acceptance criteria are given. Since the new mass window mS1 in [124.8, 230.0] GeV and the allowed ranges for kappa_H1, kappa_H2, and lambda_12 are central quantitative results, the absence of a documented scan procedure makes the claim that the entire allowed parameter space was identified difficult to verify. Please provide the scan setup or clarify the sampling method.","section":"Sec. 2.1.2, 'exhaustive' scan claim"},{"comment":"There is an internal tension between the abstract's claim that adding more singlets 'will not change this picture dramatically' and the Sec. 5 statement that 'adding more singlets would make the bounds on masses and portal couplings increasingly loose.' If the bounds become increasingly loose with N, then the picture does change quantitatively with N. The authors should either quantify the expected scaling or explicitly distinguish qualitative stability from quantitative loosening.","section":"Sec. 3.1 and Sec. 5, consistency of conclusions"}],"minor_comments":[{"comment":"The 'NE' column for sqrt(s)=14 TeV appears to use L=450 fb^-1 rather than L=3000 fb^-1: for mS1=124.8 GeV, 3.270 fb x 3000 fb^-1 = 9810, not 1471. The same pattern is repeated for the other rows. Please correct the entries and re-check any statements that depend on them.","section":"Table 3"},{"comment":"The barred couplings kappa_H1, kappa_H2, and kappa_H12 used in Figs. 9 and 10 are never explicitly defined in relation to the unbarred couplings in Eqs. (2.18)-(2.20). Please define the notation.","section":"Sec. 2.2 and Figs. 9-10"},{"comment":"The abstract states that one-singlet DM masses below about 3.5 TeV are excluded, while Sec. 1 says masses above about 4 TeV are allowed with large coupling. These numbers are not contradictory in principle, but the text should state the exact boundary and the coupling dependence to avoid confusion.","section":"Abstract and Sec. 1"},{"comment":"The captions of Figs. 15-17 contain thesis-style headings and repeated text (e.g., '4.5. Searches for the Lighter DM Particle at the LHC') that do not belong in journal figure captions. The captions should be cleaned and the 'ggh' versus 'gggh' terminology checked.","section":"Figs. 15-17 captions"},{"comment":"There is a typo: 'RS, MG and TT are are partially supported' should read 'are partially supported.'","section":"Acknowledgments"},{"comment":"The kinetic terms are written as (1/2)(partial_mu S1) partial^mu S1; the notation should be (1/2)(partial_mu S1)(partial^mu S1) for clarity. This appears in several Lagrangian displays.","section":"Eq. (2.3) and related kinetic terms"}],"recommendation":"major_revision","confidential_remarks":"The two-singlet independent-Z2 result is a solid, publishable core: the fraction-weighted direct-detection treatment is correct and the LHC benchmark study is useful. The main risk is overclaiming the N>3 stability in the abstract on the basis of a deliberately restricted three-singlet scan. I would not reject the paper; I would require the authors to either perform the missing all-heavy/all-light scans (or a four-singlet test) or to soften the N-singlet claim to match what the scans actually show. The same-Z2 section also needs a documented scan and a discussion of the mass-splitting dependence before its 'entire mass range' claim can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read for you: the two-singlet result is the real content, and it holds up. Adding a second real singlet with its own Z2 opens a window with mS1 in [124.8, 230] GeV, relic fraction around 1e-7, and mS2 in [4.3, 10] TeV, and the paper correctly treats the direct detection bound as scaled by the relic fraction. The LZ 2024 overlay showing most of that window inside the experimental uncertainty band is timely and useful. I checked the fraction-weighting and the micrOMEGAs usage; that part is sound. The one-Z2 two-singlet case is also a nice point: a small DD-coupled coupling plus large relic-driving couplings evades the classic singlet constraints.\n\nThe soft spots are concentrated in the three-singlet section and in the packaging. The claim that 'adding more singlets will not change this picture dramatically' is not established. The scan covers only two hierarchies, and the justification that no new annihilation channels open is insufficient, because the one-light-two-heavy region itself arises from shared relic fractions relaxing DD, not from new depletion. By that same mechanism an all-heavy hierarchy could shift the bounds further; the paper's own conclusion that bounds become 'increasingly loose' with N is a quantitative change that sits uneasily with the stability message. That should be rewritten as an explicit extrapolation, not a conclusion.\n\nSmaller issues: the benchmark portal couplings reach values close to the 4pi perturbativity limit, the LHC cross sections are quoted without uncertainties and with A*epsilon set to 1, and no scan data or code are released. The uploaded manuscript also contains corrupted/duplicated material (unrelated gauge-theory text in Figure 1, duplicated thesis-like passages in Section 4); the authors need to clean that up before publication.\n\nWho benefits: DM phenomenologists working on singlet extensions, and experimental colleagues who want concrete targets for HL-LHC mono-Higgs/mono-jet and next-generation DD. The two-singlet window is a defensible new observation despite the caveats, so this deserves a serious referee rather than a desk reject. I would ask for a revised version that narrows the N-singlet claim, releases the scan points, and fixes the production issues.","headline":"A solid two-singlet DM scan with a genuinely new light-heavy window; the three-singlet stability claim is softer than the abstract suggests.","tokens_in":27131,"tokens_out":1502,"would_cite":true,"duration_ms":16538,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d"],"model":"deepseek-v4-flash","headline":"Adding a second singlet with its own Z2 symmetry opens a light dark-matter window that the one-singlet model closes.","keywords":["dark matter","real scalar singlet extension","Z2 symmetry","Higgs portal","relic density","direct detection","multi-component dark matter","LHC mono-X searches"],"falsifier":"A full uniform scan of the three-singlet parameter space that finds an allowed region outside the two scanned hierarchies, or a next direct-detection limit that drops below $\\sigma_{\\rm SI}(S_1N\\to S_1N)\\,\\Omega_{S_1}/\\Omega_{\\rm DM}\\sim10^{-47}\\,{\\rm cm}^2$ for $m_{S_1}$ between 125 and 230 GeV, would falsify the paper's central picture.","tokens_in":25980,"feed_emoji":"🌌","tokens_out":8678,"duration_ms":80977,"temperature":0.7,"pith_summary":"The paper asks whether the simplest extensions of the Standard Model that produce dark matter—real scalar singlets coupled to the Higgs—can still be viable and observable. It establishes that with one singlet, only a very heavy state (above about 3.5 TeV) or a resonant state near half the Higgs mass survives relic-density and direct-detection constraints. Adding a second real singlet with an independent Z2 symmetry changes that: a light state in the window $m_{S_1}\\in[124.8,230.0]$ GeV can survive, provided it carries only a tiny relic fraction $\\Omega_{S_1}h^2\\sim10^{-7}$, while a heavy partner between about 4.3 and 10 TeV supplies the measured dark matter abundance. This matters because the light state has a large portal coupling and can be produced at the LHC, so the model becomes testable in mono-Higgs and mono-jet searches at the HL-LHC.","feed_headline":"Two dark matter particles open a light mass window near the Higgs","feed_subtitle":"A 125-230 GeV candidate with a tiny relic share evades direct detection while a 4-10 TeV partner supplies the dark matter.","key_machinery":"The mechanism carrying the argument is the relic-fraction-weighted direct-detection cross section, $\\sigma_{\\rm SI}(S_r N\\to S_r N)\\,\\Omega_{S_r}/\\Omega_{\\rm DM}$, together with the unbroken $\\mathbb{Z}_2^{(1)}\\times\\mathbb{Z}_2^{(2)}$ symmetries that make both singlets stable dark matter candidates. Because a state with a tiny relic share is allowed a much larger Higgs portal coupling, the light state can live at $m_{S_1}\\in[124.8,230.0]$ GeV while the heavy state supplies most of the relic density; the small inter-dark coupling $\\lambda_{12}$ controls heavy-to-light annihilation and sets that share. In the single-$\\mathbb{Z}_2$ variant, rotation to mass eigenstates redefines the couplings so that only one effective coupling enters direct detection while all three portal couplings contribute to the relic density.","core_discovery":"The paper's central claim is that increasing the number of real singlet fields, each protected by its own unbroken $\\mathbb{Z}_2$ symmetry, opens parameter space that the single-singlet model closes. In the two-singlet model the allowed region contains a new one-light-one-heavy configuration: $m_{S_1}\\in[124.8,230.0]$ GeV with $\\Omega_{S_1}h^2\\sim10^{-7}$ and $m_{S_2}\\in[4321.0,9977.0]$ GeV with $\\Omega_{S_2}h^2\\simeq\\Omega_{\\rm DM}h^2$. Direct detection is evaded because the bound applies to $\\sigma_{\\rm SI}(S_1N\\to S_1N)\\,\\Omega_{S_1}/\\Omega_{\\rm DM}$, and the tiny relic fraction compensates for the large portal coupling $\\kappa_{H1}\\in[4.066,9.986]$. In the three-singlet model the same logic allows two heavy states to share the relic density, weakening the usual mass and coupling bounds, and the paper argues that further singlets will only loosen these constraints. When two singlets are odd under one shared $\\mathbb{Z}_2$, mixing among dark scalars leaves only one coupling controlling direct detection, so the lightest state can populate the entire range from half the Higgs mass to the TeV scale.","pith_inferences":["The relic-fraction suppression used here is not specific to singlet scalars; any multi-component dark matter model with a light state that annihilates efficiently could open sub-TeV windows, so the mechanism provides a general template for evading direct-detection limits.","If LZ or its successor excludes the predicted $\\sigma\\times$fraction band, the two-singlet window would be closed unless the light state is even lighter or even more depleted; that outcome would push these models toward the resonant $m_h/2$ region.","The three-singlet conclusion about stability across $N$ is only as strong as the scanned hierarchies; a uniform scan, or a dedicated all-light search, would be a direct test of whether the picture really stabilizes.","For the single-$\\mathbb{Z}_2$ two-singlet model, the combination of invisible Higgs width and mono-Higgs measurements could constrain the mixing angle $\\alpha$, providing a complementary handle beyond direct detection."],"forward_implications":["The light state in the two-singlet window, with $m_{S_1}\\approx125\\text{--}230$ GeV and $\\kappa_{H1}\\gtrsim5$, has LHC production cross sections that mono-Higgs searches already approach within about one order of magnitude; the HL-LHC should be able to probe or exclude it.","The heavy state $S_2$, carrying almost all the relic density, sits at masses 4.3--10 TeV where its spin-independent cross section falls within the LZ 2024 uncertainty band; the next direct-detection exposure will likely test the whole new window.","In the three-singlet one-light-two-heavy case, two heavy states each carry part of the relic density, so the direct-detection bound on each is weakened by its fraction; this produces allowed points with heavier masses and larger portal couplings than the two-singlet model permits.","Adding further singlets with independent $\\mathbb{Z}_2$ symmetries should continue the trend: the lightest state can stay near the Higgs mass with a negligible relic fraction, while the remaining states share the observed abundance.","In the one-$\\mathbb{Z}_2$ variant, the mass eigenstates mix and only one effective portal coupling enters direct detection, so a DM candidate at any mass from $m_h/2$ to the TeV scale is allowed in principle and accessible to mono-$X$ searches."],"supporting_citations":[{"why":"It provides the observed relic density $\\Omega_{\\rm DM}h^2=0.120\\pm0.001$ that the allowed bands must match.","marker":"[11]"},{"why":"It defines the two-real-singlet setup and supplies the perturbative-unitarity and boundedness constraints used in the scans.","marker":"[14]"},{"why":"It computes the relic densities and direct-detection rates for the multi-component dark sectors.","marker":"[20]"},{"why":"It supplies the LUX-ZEPLIN spin-independent WIMP-nucleon limits that define the surviving parameter space.","marker":"[31]"},{"why":"It reports the 2024 LZ results whose uncertainty band the new allowed points occupy.","marker":"[35]"},{"why":"It generates the LHC production cross sections for mono-jet, mono-Higgs, mono-$Z$ and $b\\bar{b}$ final states.","marker":"[36, 37]"},{"why":"It provides the ATLAS mono-Higgs upper limits that the benchmark predictions most closely approach.","marker":"[40]"},{"why":"It provides the ATLAS multi-jet plus missing-energy upper limits used for the mono-jet comparison.","marker":"[39]"}],"fun_headline_variants":["Two-singlet dark matter: a light state pairs with a heavy one","Light singlet dark matter evades detection near the Higgs","Multi-singlet models open a light dark matter window","A second singlet lets dark matter hide just above the Higgs","Two dark scalars: one light and elusive, one heavy and dominant"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The broad conclusion that adding more singlets will not change this picture dramatically rests on the assumption that no new depletion channels open in unscanned regions of the three-singlet parameter space, because the scans cover only two-light-one-heavy and one-light-two-heavy hierarchies.","fun_headline_variants_meta":{"raw":{"variants":["Two-singlet dark matter: a light state pairs with a heavy one","Light singlet dark matter evades detection near the Higgs","Multi-singlet models open a light dark matter window","A second singlet lets dark matter hide just above the Higgs","Two dark scalars: one light and elusive, one heavy and dominant"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000279,"raw_usage":{"total_tokens":1749,"prompt_tokens":1131,"completion_tokens":618,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":747,"completion_tokens_details":{"reasoning_tokens":529}},"tokens_in":747,"tokens_out":618,"duration_ms":6354,"temperature":1.0,"reasoning_tokens":529,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:08:45.990988+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A full uniform scan of the three-singlet parameter space that finds an allowed region outside the two scanned hierarchies, or a next direct-detection limit that drops below $\\sigma_{\\rm SI}(S_1N\\to S_1N)\\,\\Omega_{S_1}/\\Omega_{\\rm DM}\\sim10^{-47}\\,{\\rm cm}^2$ for $m_{S_1}$ between 125 and 230 GeV, would falsify the paper's central picture.","supporting_citations":[{"cited_title":"New Dark Matter Search Results from the LUX-ZEPLIN (LZ) Ex- periment","cited_arxiv_id":null,"evidence_quote":"It reports the 2024 LZ results whose uncertainty band the new allowed points occupy."},{"cited_title":"Search for dark matter produced in association with a Standard Model Higgs boson decaying into b-quarks using the full Run 2 dataset from the ATLAS detector","cited_arxiv_id":null,"evidence_quote":"It provides the ATLAS mono-Higgs upper limits that the benchmark predictions most closely approach."}],"review_version":1}