{"id":"2057b14b-d049-4913-8a3e-0b09399ed82a","arxiv_id":"1908.08554","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The two-real-singlet extension of the Standard Model can produce sizable rates for asymmetric and cascade Higgs-to-Higgs decays at the LHC, and the paper supplies six benchmark planes for those signatures.","lead":"This paper maps out the LHC signatures of a model with two extra real scalar singlet fields, one of which mixes into the observed 125 GeV Higgs boson. It identifies asymmetric and cascade decays to multiple Higgs bosons as new, potentially visible signatures and provides six benchmark scenarios to guide experimental searches.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"NWA and peak-signal-only use for asymmetric/cascade benchmarks should be checked: box-diagram interference and finite-width effects can shift cross sections around thresholds","rationale":"The reader's weakest_assumption was the narrow-width approximation, kappa-squared rescaling, and neglected interference/off-shell effects, including near-degenerate scalars and Gamma/M ~ 14-18%. My stress test agrees and sharpens the concern with specific quotes: Eq. (40), Sec. IV B's statement on interference, the BP5 near-degeneracy (M1 up to 124 GeV with h125 = h2), BP6's Gamma3/M3 ~ 14%, and the sharp h2->h1h1 threshold in BP1. The paper provides independent support via constraints from HiggsBounds/HiggsSignals and explicit benchmark points, and the qualitative existence of sizable Higgs-to-Higgs decays is not in doubt. However, the quantitative claim of specific signal rates for experimental searches depends on the NWA factorization; the cited Ref. [119] work on signal-signal interference in di-Higgs production shows that near-degenerate configurations can give order-one effects, and the paper does not quantify this for the benchmarks. Therefore the verdict CONDITIONAL is appropriate: the paper should be accepted only if the authors add a finite-width/interference check for the benchmark planes or clearly flag the affected regions. I would not REJECT because the central qualitative claim is supported and the paper itself flags the approximation in places (e.g. near BP6 threshold).","tokens_in":27106,"tokens_out":1833,"duration_ms":17538,"concrete_test":"Recompute the BP2 (h3->h125h1) and BP6 (h3->h2h2) benchmark planes keeping finite widths and including the full s-channel Breit-Wigner propagator and the h1h2/h2h2 box-diagram amplitudes (e.g. with MadGraph5_aMC@NLO or a dedicated calculation), instead of Eq. (40). Also compute the interference between h2 and h125 contributions in BP5 for M1 near 124 GeV by evaluating the full g->h_i->h1h1 amplitude with physical propagators and widths. If the inclusive pp->h_a->h_b h_c cross sections change by more than ~30% in any region where the paper quotes a maximal or near-threshold rate, the benchmark tables and Figs. 6-15 need revision.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that asymmetric and cascade Higgs-to-Higgs decays reach sizable rates, e.g. O(3 pb) in BP1, O(0.6 pb) in BP2, O(0.3 pb) in BP3, and O(14 fb) for the BP6 double-cascade, all based on a narrow-width factorization: sigma(pp->ha->hbhc) = kappa_a^2 sigma_SM(Ma) BR(ha->hbhc), Eq. (40). This is only the s-channel single-resonance contribution. The paper's own Sec. IV B notes that box diagrams and signal-signal interference can significantly affect di-Higgs production (Ref. [119]) and asserts without a dedicated check that such configurations play no important role for most scenarios. The benchmarks include regions where this is not automatically safe: BP5 allows M1 up to 124 GeV alongside h125 = h2 at 125 GeV, a near-degenerate configuration of exactly the type shown to produce large interference in the two-Higgs-doublet studies the paper cites; BP6 reaches Gamma3/M3 ~ 14% (Fig. 14), and BP1's h2->h1h1 threshold at M2 = 2M1 has an extremely sharp transition (a step in BR of ~100%) where the NWA and even the tree-level width formula are unreliable. The quoted multi-pb signal rates in BP4, and the 14 fb four-Higgs rate in BP6 near threshold, could shift by order-one factors if off-shell and interference effects are included. The qualitative conclusion that sizable rates exist is likely robust, but the specific benchmark cross sections and even some kinematic boundaries (e.g. which masses are excluded by h125 coupling constraints) could change.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the two-real-scalar-singlet extension of the Standard Model (TRSM), in which an imposed Z2⊗Z2′ symmetry is spontaneously broken by the singlet vevs, giving three CP-even neutral scalars h1,2,3, one of which is identified with the 125 GeV Higgs boson. The authors parametrize the scalar potential in terms of physical masses, mixing angles, and vevs, and derive the resulting couplings, partial widths, and LHC signal rates. They apply theoretical constraints (boundedness, perturbative unitarity), electroweak precision constraints, and the public codes HiggsBounds and HiggsSignals to map the allowed parameter space. The central phenomenological claim is that, in addition to standard direct decays to SM particles, the model generically produces sizable Higgs-to-Higgs decays, including asymmetric decays h3→h1h2, symmetric decays h3→h1h1, h2→h1h1, and cascades such as h3→h2h2→h1h1h1h1. Six two-dimensional benchmark planes (BP1–BP6) are proposed to target these signatures, with cross sections ranging from tens of picobarns for light states to about 14 fb for the four-Higgs cascade final state near threshold.","tokens_in":27395,"tokens_out":3481,"duration_ms":42283,"significance":"If the quantitative rates stand, the paper fills a genuine gap: LHC searches for Higgs-to-Higgs decays have so far concentrated on h125h125 and h125→light+light, while asymmetric decays and cascades involving only non-SM scalars have received little experimental attention. The paper's strengths are its transparent physical parametrization, the explicit coupling and width formulas, the use of state-of-the-art public constraints (HiggsBounds, HiggsSignals, electroweak precision data), and the concrete, reproducible six benchmark parameter sets that experimental collaborations can adopt directly. The qualitative conclusion that these signatures can dominate over direct SM decays is robust, and the benchmark definitions are a useful community resource even if some absolute rates require further refinement.","major_comments":[{"comment":"The central numerical predictions are computed in the narrow-width approximation with the factorized formula σ(pp→ha→hbhc) = κa^2 σSM(Ma) BR(ha→hbhc), i.e. only the s-channel single-resonance contribution is included. The paper notes in Sec. IV B that box diagrams and signal-signal interference can significantly affect di-Higgs production, but asserts without a dedicated check that such configurations \"play no important role for most of the scenarios.\" This assertion is load-bearing because the benchmark planes include exactly the configurations where interference is known to be important: BP5 allows M1 up to 124 GeV while h125≡h2 has M2=125.09 GeV, and BP1 has a sharp threshold M2=2M1 across which BR(h2→h1h1) jumps from ~0 to ~100%. I request a quantitative estimate of off-shell and interference effects for at least the representative benchmark points, or a clear restriction of the claimed benchmark rates to regions where the NWA is reliable.","section":"Sec. IV B and Eq. (40)"},{"comment":"The narrow-width approximation is violated in BP6, where the total width of h3 reaches Γ3/M3 ~ 14% near the unitarity bound. The paper itself advises that \"it may be important to include finite width effects in experimental analyses of this scenario,\" yet the quoted benchmark cross sections, including the ~14 fb four-Higgs cascade rate near threshold, are obtained under the NWA. Since the cross section is proportional to the on-shell branching ratio divided by the width, a 14% width can shift rates by order-one factors and distort line shapes near thresholds. I ask the authors to provide finite-width corrected estimates, or at least to state the expected size of the correction for the quoted benchmark rates.","section":"Sec. V F and Fig. 14"},{"comment":"The Higgs-to-Higgs partial widths are computed at leading order in the effective trilinear couplings, while the SM-like partial widths are taken from higher-order calculations. The branching ratios in Eqs. (26)–(28) therefore mix different perturbative orders. This is a standard limitation for benchmark studies, but it directly affects the quoted BR(h125→NP) limits and the benchmark BRs, especially in regions where a small change in a partial width can change the dominant decay mode. I ask the authors to state this limitation explicitly in Sec. II B or Sec. V, and, if feasible, to estimate the size of the leading QCD/electroweak corrections to the scalar trilinear couplings and widths.","section":"Eq. (22) and Sec. II B"}],"minor_comments":[{"comment":"In the paragraph after Fig. 14 there is a typo: \"the total width of of h3\" should read \"the total width of h3\".","section":"Sec. V F"},{"comment":"The right panel caption says the color scale shows Γa/Ma as a function of Ma and BR(ha→NP), but the plotted quantity is not fully clear from the caption alone; please clarify the color-coding and the ordering of the points.","section":"Fig. 4 caption"},{"comment":"The notation F^4_SM, F^6_SM, and F^8_SM is used in Tables III and IV before it is explicitly defined in the text of Sec. V; please define the notation at first use in Sec. IV B.","section":"Tables III and IV"},{"comment":"In BP5, the statement that the h1h1 state is identical to that of BP4 is correct only because h1 always decays like a SM Higgs of the same mass; note this explicitly in the text to avoid confusion, since the two benchmarks have different h125 assignments.","section":"Sec. V E and Fig. 13"}],"recommendation":"major_revision","confidential_remarks":"This is a useful and well-executed phenomenological paper with clearly defined benchmark scenarios. My main concern is the NWA-based quantitative rates in a few benchmark regions where finite-width and interference effects are likely to be non-negligible; however, these effects are localized and can, in principle, be addressed with additional calculations or by restricting the claims. I do not think rejection is warranted. The paper is suitable in scope for the journal after substantial revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The TRSM is the natural next step after the single-singlet and CxSM studies, and this is the first paper to work out its collider phenomenology in a systematic way. The genuinely new content is the catalog of asymmetric and cascade Higgs-to-Higgs decays — h3→h1h2, h3→h125h2, h2→h1h1, h3→h2h2→h1h1h1h1 — with six benchmark planes tailored to each. That is concrete: the rates, from ~100 pb for light scalars down to ~14 fb for the double cascade, are presented in a form experimentalists can translate into search designs. The constraints are applied with public tools, and the parametrization in terms of physical masses, mixing angles, and vevs is clean. I think the central qualitative claim holds: these exotic Higgs-to-Higgs modes can dominate direct SM decays in sizable regions of the allowed parameter space.\n\nThe soft spots are the ones the stress-test flags. The whole benchmark numerics rely on the narrow-width approximation, Eq. (40), and the authors note in Sec. IV B that signal-signal interference can affect di-Higgs production but assert it plays no important role for most scenarios without a dedicated check. That is the weakest point. BP5 allows M1 within 1 GeV of h125, a near-degenerate configuration where overlapping-resonance interference is known to be large in the 2HDM studies they cite. BP6 reaches Γ3/M3 ~ 14%, where NWA is not reliable. The quoted cross sections in those corners could shift by order-one factors. The qualitative message survives, but the benchmark numbers should be treated as approximate; the paper should estimate the finite-width/interference corrections or at least clearly flag which regions of the planes are affected. The leading-order decay widths and the private ScannerS update are secondary concerns — not fatal, but consistent with the paper being a benchmark-oriented study rather than a high-precision prediction.\n\nTo the paper's credit, it is explicit about these limitations, and the benchmark parameters are given in enough detail that others can check the claims. This deserves a serious referee, with the main request being an honest treatment — or an explicit caveat — of NWA reliability in the affected regions. I'd bring it to a reading group focused on new Higgs signatures, and I'd cite it if I needed a TRSM reference. Verdict: accept for peer review, conditional on those caveats.","headline":"First systematic LHC pheno study of the Z2 x Z2' two-real-singlet extension; the asymmetry/cascade signatures are new and the benchmark planes are useful, though the NWA-based rates need caveats near thresholds and broad widths.","tokens_in":27969,"tokens_out":1806,"would_cite":true,"duration_ms":21025,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In a minimal two-singlet extension of the Standard Model, the three Higgs bosons can decay into one another at rates that dominate their decays to ordinary particles, producing three- and four-Higgs final states at the LHC.","keywords":["two-real-singlet model","Higgs-to-Higgs decays","asymmetric scalar decays","cascade Higgs decays","scalar singlet extension","LHC phenomenology","benchmark scenarios","resonant di-Higgs production"],"falsifier":"A dedicated LHC search for $pp\\to h_3\\to h_1h_2$ in the $b\\bar{b}b\\bar{b}$ final state across the BP2 plane ($M_3$ between 150 and 500 GeV, $M_1$ up to about 125 GeV) would confirm or exclude the predicted signal rates between roughly 50 fb and 0.6 pb, directly testing whether asymmetric Higgs-to-Higgs decays dominate as claimed. In BP6, a measurement of the $h_3$ line shape for $M_3\\sim 400$--600 GeV would show whether the predicted $\\Gamma_3/M_3$ up to about 14% is physically present.","tokens_in":127,"feed_emoji":"🎯","tokens_out":14030,"duration_ms":183431,"temperature":0.7,"pith_summary":"This paper argues that in a minimal extension of the Standard Model where two real scalar singlet fields mix with the ordinary Higgs doublet, the three resulting Higgs bosons can decay into one another at rates that often dominate their direct decays to Standard Model particles. That includes asymmetric decays, in which the produced scalar turns into two different lighter scalars, and cascade decays that ultimately give three- or four-Higgs final states. The authors determine which parameter choices survive current theoretical and experimental constraints and condense the results into six two-dimensional benchmark scenarios with near-maximal signal rates for each signature. A curious reader should care because most of these signatures have not been searched for at the LHC, so the model supplies concrete targets beyond the standard pair-production of two 125 GeV Higgs bosons.","feed_headline":"Two extra scalars make Higgs-to-Higgs decays dominate at the LHC","feed_subtitle":"Six benchmark planes map unsearched three- and four-Higgs final states, with rates from picobarns to tens of fb.","key_machinery":"The machinery is the scalar spectrum of the model: three neutral CP-even Higgs states obtained by mixing the SM doublet with two singlet fields, with all couplings of a given mass eigenstate to SM particles rescaled by a single factor $\\kappa_a = R_{a1}$ and obeying the sum rule $\\sum_a \\kappa_a^2 = 1$. The triple-scalar couplings that drive Higgs-to-Higgs decays are written directly in terms of the three masses, three mixing angles, and three vacuum expectation values, and the tree-level width for $h_a\\to h_bh_c$ is given by the standard phase-space formula depending on those couplings. Production is then factorized through a narrow-width approximation, $\\sigma(pp\\to h_a) = \\kappa_a^2 \\sigma_{\\rm SM}(M_a)$, with branching ratios built from the rescaled SM widths and the new scalar decay widths. That factorization is what lets the authors convert SM Higgs predictions into complete, surveyable collider phenomenology for all three states.","core_discovery":"The central claim is that the two-real-singlet model with a spontaneously broken $\\mathbb{Z}_2\\otimes\\mathbb{Z}_2'$ symmetry contains three CP-even Higgs mass eigenstates, one of which is the observed 125 GeV boson, and that resonant Higgs-to-Higgs decays among them can be sizable, easily dominating the direct decays to SM states. The paper demonstrates branching ratios up to roughly 50--90% for channels such as $h_3\\to h_1h_2$, $h_3\\to h_{125}h_2$, and $h_3\\to h_1h_1$, together with cascade chains like $h_3\\to h_1h_2$ followed by $h_2\\to h_1h_1$, or $h_3\\to h_2h_2$ followed by $h_2\\to h_{125}h_{125}$, producing three- and four-Higgs final states. Benchmark signal rates span from tens of picobarns for light scalars down to roughly 14 fb near threshold for the four-Higgs final state. The paper also finds that current LHC searches for $h_{125}h_{125}$ and for $h_{125}\\to h_ah_a$ are not sensitive to most of these signatures, so the benchmark planes are largely untested.","pith_inferences":["The same asymmetric and cascade patterns should appear in other models with three or more Higgs bosons and comparable triple-scalar couplings, so model-independent limits on these planes would constrain a broader class of singlet and multi-Higgs extensions, not only this one.","The widest benchmark resonances reach $\\Gamma_3/M_3\\sim 14\\%$, where the narrow-width approximation is stressed; finite-width and signal-signal interference effects could shift the quoted rates, and a line-shape measurement near threshold would settle how much the predictions move.","Because the singlet vacuum expectation values enter the triple-scalar couplings but are only weakly constrained by current data, a future measurement of any Higgs-to-Higgs rate would give a direct, novel handle on the vacuum structure of the singlet sector, which precision electroweak fits barely probe."],"forward_implications":["Resonant LHC searches should target asymmetric decays such as $h_3\\to h_1h_2$ and $h_3\\to h_{125}h_2$, which reach branching ratios of 40--55% and cross sections around 0.3--0.6 pb in the benchmarks, not just the $h_{125}h_{125}$ channel.","Cascade decays produce three- and four-Higgs final states; the double cascade $h_3\\to h_2h_2\\to h_{125}h_{125}h_{125}h_{125}$ gives about 14 fb near threshold, with $b$-rich final states that should be visible with Run-II data.","The six benchmark planes BP1--BP6 each maximize one novel signature while passing all applied constraints, so experimental results in those planes can be compared directly with model predictions or recast model-independently.","Because the sum rule forces the other scalars' couplings to be small whenever the 125 GeV state is SM-like, Higgs-to-Higgs decays frequently dominate over direct SM decays, so conventional single-resonance searches can miss the model entirely."],"supporting_citations":[{"why":"Supplies the SM Higgs branching ratios and total widths used as the baseline for the $\\kappa_a^2$ rescaling of every scalar state.","marker":"[77]"},{"why":"Provides the NNLO+NNLL SM Higgs production cross sections used for scalars with masses different from 125 GeV.","marker":"[114]"},{"why":"Provides the N3LO gluon-fusion production cross section used for the 125 GeV state.","marker":"[115]"},{"why":"Implements the measured 125 GeV signal-rate constraints used to restrict the viable parameter space.","marker":"[104–107]"},{"why":"Implements existing LEP and LHC searches for additional Higgs bosons used to exclude regions of the benchmark planes.","marker":"[98–103]"},{"why":"Documents signal-signal interference effects in di-Higgs production, which the authors argue play no important role for most of their scenarios.","marker":"[119]"},{"why":"A first LHC search for the $h_3\\to h_2h_2\\to W^+W^-W^+W^-$ signature that already constrains a small region of BP6.","marker":"[61]"},{"why":"A resonant double-Higgs production search whose limit begins to constrain the $M_2\\sim 125$ GeV regions of BP5 and BP6.","marker":"[66]"},{"why":"A search for $pp\\to h_a\\to h_{125}h_{125}$ that provides the current direct experimental bound in the 380--550 GeV resonance mass range.","marker":"[69]"}],"fun_headline_variants":["Two extra scalars drive cascade decays to 3–4 Higgs states","Higgs-to-Higgs decays dominate in two-singlet extension","New scalars produce triple and quadruple Higgs final states","Six benchmark planes expose untested multi-Higgs final states","Higgs-to-Higgs branching up to 90% in two-singlet model"],"cache_read_input_tokens":29952,"weakest_assumption_plain":"All quoted rates rely on treating each scalar as a narrow resonance whose production cross section is exactly the SM value times $\\kappa_a^2$ and whose production and decay factorize, with interference and off-shell effects neglected; for the widest benchmark states $\\Gamma/M$ reaches about 14--18%, where that treatment is not obviously reliable.","fun_headline_variants_meta":{"raw":{"variants":["Two extra scalars drive cascade decays to 3–4 Higgs states","Higgs-to-Higgs decays dominate in two-singlet extension","New scalars produce triple and quadruple Higgs final states","Six benchmark planes expose untested multi-Higgs final states","Higgs-to-Higgs branching up to 90% in two-singlet model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000945,"raw_usage":{"total_tokens":4088,"prompt_tokens":1047,"completion_tokens":3041,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":663,"completion_tokens_details":{"reasoning_tokens":2946}},"tokens_in":663,"tokens_out":3041,"duration_ms":24608,"temperature":1.0,"reasoning_tokens":2946,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:36:57.079153+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A dedicated LHC search for $pp\\to h_3\\to h_1h_2$ in the $b\\bar{b}b\\bar{b}$ final state across the BP2 plane ($M_3$ between 150 and 500 GeV, $M_1$ up to about 125 GeV) would confirm or exclude the predicted signal rates between roughly 50 fb and 0.6 pb, directly testing whether asymmetric Higgs-to-Higgs decays dominate as claimed. In BP6, a measurement of the $h_3$ line shape for $M_3\\sim 400$--600 GeV would show whether the predicted $\\Gamma_3/M_3$ up to about 14% is physically present.","supporting_citations":[],"review_version":1}