{"id":"14fdd0f4-8078-4318-a44a-7f6aec14f204","arxiv_id":"2504.19886","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"A new dark matter model makes a pseudo-Goldstone boson stable with a residual Z3 symmetry and obtains the observed relic abundance from annihilation plus semi-annihilation.","lead":"Three complex scalar fields with a dark U(1) gauge symmetry produce a pseudo-Goldstone dark matter particle stabilized by a residual Z3 symmetry. The model permits new semi-annihilation processes and a viable relic abundance, but the boosted dark matter signal it predicts is too small to detect.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Global-minimum proof for the equal-VEV vacuum is missing; the Appendix A Z3-breaking stationary point could destabilize the DM candidate if it lies deeper.","rationale":"The reader's weakest assumption matches the most load-bearing concern I can identify. The equal-VEV vacuum is the foundation on which the Z3-stabilized pNGB dark matter candidate rests; if this vacuum is metastable rather than the global minimum, the dark matter candidate is not actually stable and the relic-density, semi-annihilation, and direct-detection statements all lose their basis. Appendix A acknowledges a competing stationary point but dismisses it by symmetry rather than by comparing potential depths. This is a genuine gap in the argument, not a value judgment about the model. For the quoted benchmarks, mΣ=3mDM makes λS−λ′S positive, so the quartic sector favors equal magnitudes and the soft term is minimized by equal phases, making the equal-VEV vacuum very likely the global minimum; however, the paper gives no analytic or numerical demonstration. A secondary observation is that the printed a0 matrix in eq. (3.2) has a diagonal entry λS at position (5,5), while the S3S3* self-scattering amplitude should be 2λS by S(3) symmetry; the stated unitarity inequalities appear to correspond to the corrected entry, so this is likely a typographical slip rather than a fatal flaw, but it should be fixed. The conditional verdict is appropriate: the model is coherent and the central claim is probably correct, but the missing vacuum-depth check should be supplied before the paper can be accepted without qualification.","tokens_in":15601,"tokens_out":29125,"duration_ms":304440,"concrete_test":"For the Fig. 3 benchmark (sinθ=0.1, sinϵ=10^−4, m2=300 GeV, mΣ=3mDM, mZ′=200 GeV) and several mDM values, numerically minimize the full scalar potential including complex phases and compare the depth of the equal-VEV vacuum with every stationary point found from Appendix A. Also test the limiting cases mΣ=mDM and λS=λ′S, where the quartic preference for equal magnitudes weakens. If any Z3-breaking stationary point is deeper for a point on the relic curve, the central claim fails; otherwise the paper should add a short proof or a numerical scan demonstrating the equal-VEV vacuum is the global minimum.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central viability claim assumes the scalar potential is minimized at ⟨S1⟩=⟨S2⟩=⟨S3⟩=vs/√6, the configuration that preserves the stabilizing Z3. Appendix A explicitly finds a second stationary point with vs1≠0, vs2=0, vs3=−vs1 that satisfies the minimization equations, and it is discarded only because it breaks the residual Z3. No comparison of potential depths is presented. This is load-bearing because the stability of aω, the relic-density calculation, and the direct-detection suppression all presuppose that this equal-VEV vacuum is the realized global minimum. The concern is concrete: the m12^2 soft term gives a positive energy contribution at the competing point, so the equal vacuum is very likely lower for the benchmark choices (mΣ=3mDM implies λS−λ′S>0), but the paper provides no inequality or numerical check showing the competing point is never deeper. If the Z3-breaking stationary point were the true minimum, the DM candidate would not be stable and the paper's phenomenological conclusions would not follow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a dark matter model with three complex scalar fields, a dark U(1)_V gauge symmetry, a softly broken global U(1)_A symmetry, and an exact S(3)/Z3 structure. It derives the scalar mass spectrum, identifies a complex pNGB a_omega as a stable DM candidate, and computes constraints from perturbative unitarity and Higgs invisible decays. Relic abundance contours are obtained with micrOMEGAs, emphasizing new semi-annihilation channels a_omega a_omega -> a*_omega Z' and a_omega a_omega -> Sigma*_omega h2, and the paper closes with an estimate of boosted DM elastic scattering. The central claim is that a viable thermal parameter space exists for this semi-annihilating pNGB DM with naturally suppressed direct detection.","tokens_in":15806,"tokens_out":32198,"duration_ms":339231,"significance":"If the vacuum issue is resolved, the model is a useful minimal realization of semi-annihilating pNGB DM that goes beyond the existing Z2 two-scalar construction: it explicitly derives the Z3 cubic interactions, shows their effect on the relic abundance, and gives a concrete negative estimate of BDM detection prospects. The use of FeynRules and micrOMEGAs makes the numerical results reproducible, and the paper appropriately treats the v/v_s curve as a fitted contour rather than a parameter-free prediction. The missing proof that the equal-VEV vacuum is the global minimum is a clear and likely fixable gap rather than a contradiction in the derivations.","major_comments":[{"comment":"The equal-VEV vacuum <S1>=<S2>=<S3>=vs/sqrt(6) is assumed in Eq. (2.7), but Appendix A explicitly finds a competing stationary point with vs1 != 0, vs2 = 0, vs3 = -vs1 that satisfies the minimization equations (A.2)-(A.5). This point is discarded only because it breaks the residual Z3, and no comparison of potential depths is presented. Since the stability of a_omega, the relic calculation, and the direct-detection cancellation all presuppose that the equal-VEV vacuum is the realized global minimum, the paper should prove, or impose inequalities ensuring, V(equal) <= V(competing) over the entire scanned parameter region, including the m12^2 dependence visible in Appendix A.","section":"Sec. 2.2 and Appendix A"}],"minor_comments":[{"comment":"The decay written as h1 -> a_omega a_omega violates Z3 charge conservation, because a_omega carries charge omega so a_omega a_omega has charge omega^2 while h1 is neutral. The correct final state is a_omega a*_omega, and the width for a distinguishable complex-scalar pair is Gamma = kappa^2 beta/(16 pi m1), a factor 2 larger than Eq. (3.7). This should be corrected and the red exclusion regions in Figs. 3-5 regenerated.","section":"Sec. 3.2, Eq. (3.7)"},{"comment":"The sentence \"for mZ' = 1.1 mDM, the forbidden Z'Z' channel remains open\" is internally inconsistent: for mZ' = 1.1 mDM the channel a_omega a*_omega -> Z'Z' is kinematically closed (2mZ' > 2mDM), and the semi-annihilation channel is also closed because mZ' > mDM. Please clarify the intended kinematics and identify which curve in Fig. 4 corresponds to this case.","section":"Sec. 4.2 and Fig. 4"},{"comment":"Because the U(1)_V current couples a_omega only off-diagonally to Sigma_omega, there is no tree-level a_omega-a_omega-Z' vertex. This fact is implicit in Eq. (4.1) and is the reason the Z' portal does not spoil the direct-detection suppression; it should be stated explicitly.","section":"Sec. 4.1, Eq. (4.1)"},{"comment":"There are small presentation issues: the introduction says \"charged under three dark U(1)_V gauge symmetry\" instead of \"a dark U(1)_V gauge symmetry\", and the text refers to a dashed orange curve in Fig. 4 while the caption says dashed yellow. These should be harmonized.","section":"Sec. 2.1 and Fig. 4"},{"comment":"The axes are not labeled in the displayed version of Fig. 3; please state explicitly which axis is mDM and which is v/v_s, and note that mZ' and mSigma are fixed as functions of mDM in the benchmark scan.","section":"Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The global-minimum issue is the main technical gap and I expect it can be fixed with an analytic comparison of the two stationary points. The symmetry-factor error in the invisible width is straightforward to correct but does require rerunning the constraint plots. The model itself is sound and within the scope of JHEP."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real extension, not a recycled Z2 pNGB model. Three complex scalars with a residual Z3 and the κ2 cubic terms give semi-annihilation channels that are genuinely absent in the two-scalar model of Ref. [14]. The mass spectrum, cubic vertices, and gauge diagonalization are internally consistent, and the authors are honest that the boosted-DM flux they can produce lies below experimental reach. That last point matters: they don't oversell the BDM motivation.\n\nWhat the paper does well: the relic-density comparison with the Z2 case (Figs. 3–5) makes the semi-annihilation contribution concrete, including the kink at m_DM = m_Z' and the interesting forbidden-channel behavior for m_Σ = 1.5 m_DM. The parameter space satisfying relic abundance, perturbative unitarity, and Higgs invisible decay is plausible. The perturbative-unitarity matrix and the VEV analysis in Appendix A are presented in enough detail to follow the algebra. The citation pattern is appropriate; it builds on the pNGB program without inflating the novelty.\n\nThe largest substantive gap is vacuum stability. Appendix A finds a stationary point (v_s1, 0, −v_s1) that satisfies all minimization equations and discards it because it breaks the residual Z3. That is not a reason; if it were the global minimum, the DM candidate would not be stable. For the benchmark choices the energy comparison almost certainly favors the equal-VEV vacuum: the m_12^2 soft term is positive at the competing point, and m_Σ = 3 m_DM forces λ_S > λ'_S. But the paper should show this explicitly, not assume it. This is a patchable omission, not a fatal one.\n\nMinor reproducibility issue: no model files, scripts, or benchmark tables are provided, so the relic-density curves can't be reproduced without contacting the authors. For a numerical-scan paper, that is worth requesting at review.\n\nWho is this for? People working on pNGB dark matter or semi-annihilation phenomenology. It is a genuine extension with a clear qualitative signature, and the physics looks sound. I would send it to a serious referee, with the global-minimum proof and numerical reproducibility as the main requests.","headline":"Legitimate new Z3 pNGB dark matter model with semi-annihilation; the missing global-minimum proof is patchable and should not block review.","tokens_in":16379,"tokens_out":3944,"would_cite":true,"duration_ms":39928,"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":"Three complex scalars can produce a Z3-stabilized pseudo-Nambu-Goldstone dark matter candidate, and new semi-annihilation channels help set its relic abundance.","keywords":["pseudo-Nambu-Goldstone boson dark matter","Z3 symmetry","semi-annihilation","dark U(1) gauge symmetry","relic abundance","boosted dark matter","Higgs invisible decay","perturbative unitarity"],"falsifier":"Evaluate the scalar potential at the competing stationary point $v_{s1} \\neq 0$, $v_{s2} = 0$, $v_{s3} = -v_{s1}$ from Appendix A for a benchmark point and compare it with the equal-VEV vacuum; if the Z3-breaking point is deeper, the $a_\\omega$ state is not the ground state and the model's dark-matter conclusion fails.","tokens_in":15359,"feed_emoji":"🌌","tokens_out":8279,"duration_ms":74284,"temperature":0.7,"pith_summary":"The paper proposes a dark matter model in which the dark matter candidate is a pseudo-Nambu-Goldstone boson (pNGB) made stable by a residual Z3 symmetry. The setup uses three Standard-Model-singlet complex scalars charged under a dark U(1)_V gauge symmetry, with a softly broken global U(1)_A; this extends the usual Z2 pNGB construction and opens up semi-annihilation processes. The central claim is that a consistent parameter space exists in which the pNGB a_omega reproduces the observed relic abundance through both ordinary annihilation and the new semi-annihilation channels, while direct detection is naturally suppressed by the pNGB cancellation mechanism. The paper checks perturbative unitarity and Higgs invisible decay constraints, and it evaluates the boosted-dark-matter flux, finding it too small for current detectors in the minimal setup.","feed_headline":"Z3-stabilized pseudo-Goldstone boson explains dark matter","feed_subtitle":"New semi-annihilation channels set the relic density while direct detection stays naturally suppressed.","key_machinery":"The central object is the Z3 clock symmetry in the Higgs basis: after rotating the three scalars by the matrix $R$, the fields $(\\Sigma_1, \\Sigma_2, \\Sigma_3)$ transform under $T_3 = \\mathrm{diag}(1, \\omega, \\omega^2)$, so $\\Sigma_1$ is uncharged and $\\Sigma_2, \\Sigma_3$ carry opposite Z3 charges. The dark matter candidate $a_\\omega$ is the dark-CP-odd combination of $\\Sigma_2$ and $\\Sigma_3^*$, a pseudo-Nambu-Goldstone boson: a scalar whose mass is protected by the soft breaking of a global $\\mathrm{U}(1)_A$ and whose derivative interactions suppress low-energy scattering. The argument is carried by the cubic interaction $\\kappa_2 (\\Sigma_\\omega^3 + \\Sigma_\\omega^{*3} - \\Sigma_\\omega a_\\omega^2 - \\Sigma_\\omega^* a_\\omega^{*2})$ with $\\kappa_2 = (m_\\Sigma^2 - m_\\mathrm{DM}^2)/(2 v_s)$, which generates the semi-annihilation channels, together with the $h_1$-$h_2$ mixing that produces the cancellation of the dark-matter-nucleon amplitude at zero momentum transfer.","core_discovery":"On the paper's own terms, the central discovery is that a pNGB dark matter candidate can be stabilized by a Z3 rather than Z2 symmetry using three complex scalars, without losing the direct-detection protection that makes pNGB models attractive. The mass spectrum contains a complex pNGB $a_\\omega$ with $m_a^2 = m_{12}^2$, essentially independent of the quartic couplings, and a Z3-singlet scalar sector that mixes with the SM Higgs. Because the Z3 symmetry permits cubic scalar interactions, the dark matter can undergo semi-annihilation processes $a_\\omega a_\\omega \\to a_\\omega^* Z'$ and $a_\\omega a_\\omega \\to \\Sigma_\\omega^* h_2$, which alter the thermal freeze-out and are forbidden in Z2 models. The paper establishes that for benchmark parameters ($\\sin\\theta = 0.1$, $\\sin\\epsilon = 10^{-4}$, $m_2 = 300$ GeV, $m_\\Sigma = 3 m_\\mathrm{DM}$, $m_{Z'} = 200$ GeV) a relic-abundance curve consistent with $\\Omega_\\mathrm{DM} h^2 = 0.12 \\pm 0.001$ exists, and that perturbative unitarity restricts $m_\\mathrm{DM}$ to about 2.5 TeV.","pith_inferences":["I infer that the model's most distinctive observable is not the elastic scattering of boosted dark matter but the production of a $Z'$ boson plus missing energy from $a_\\omega a_\\omega \\to a_\\omega^* Z'$, since the benchmark elastic cross section is far below reach.","One check the paper leaves open is a direct comparison of the potential depths at the equal-VEV vacuum and the competing Z3-breaking stationary point; until that is done, the stability of $a_\\omega$ rests on an unverified global-minimum assumption.","Because the paper notes that $\\lambda_S = \\lambda'_S$ makes the charged scalars degenerate and could lead to multi-component dark matter, the same three-scalar framework could be pushed toward a two-component scenario; that direction is not developed here."],"forward_implications":["When $m_\\mathrm{DM} > m_{Z'}$, the semi-annihilation process $a_\\omega a_\\omega \\to a_\\omega^* Z'$ enhances the annihilation cross section and lowers the relic abundance relative to the Z2 model.","For $m_{Z'}$ close to $m_\\mathrm{DM}$ (for example $1.1\\, m_\\mathrm{DM}$), the relic-density curve drops sharply because both the $Z'Z'$ annihilation channel and the semi-annihilation channel stay kinematically open.","With $m_\\Sigma = 1.5\\, m_\\mathrm{DM}$, a forbidden semi-annihilation-like channel $a_\\omega a_\\omega \\to \\Sigma_\\omega^* h_2$ becomes active and produces qualitatively different relic-abundance behavior than a Z2 model with the same spectrum.","The benchmark parameter choices leave a viable region below $m_\\mathrm{DM} \\simeq 2.5$ TeV after perturbative unitarity and Higgs invisible decay constraints, with resonance dips at $m_\\mathrm{DM} = m_1/2$ and $m_2/2$ and a kink at $m_\\mathrm{DM} = m_{Z'}$.","The boosted-dark-matter elastic cross section in the minimal setup is estimated at $\\sim 10^{-54}\\,\\mathrm{cm}^2$ for a benchmark point, several orders below current and near-future detector sensitivities, so no observable boosted-dark-matter signal is expected without additional boost sources."],"supporting_citations":[{"why":"Establishes the cancellation mechanism for dark-matter-nucleon scattering in pNGB models, the direct-detection suppression this model inherits.","marker":"[4]"},{"why":"Supplies the observed dark-matter relic density that the parameter scan must reproduce.","marker":"[5]"},{"why":"The two-complex-scalar Z2 pNGB model that this work extends by adding a third scalar and Z3 semi-annihilation channels.","marker":"[14]"},{"why":"Introduces the boosted-dark-matter-from-semi-annihilation scenario whose detectability motivates the phenomenological analysis.","marker":"[21]"},{"why":"Computes boosted dark matter from semi-annihilations in the galactic center, the signal the paper evaluates and finds too small.","marker":"[24]"},{"why":"Provides the perturbative-unitarity conditions used to bound the scalar quartic couplings and the dark gauge coupling.","marker":"[26]"},{"why":"Supplies one of the experimental upper bounds on Higgs invisible decay used to exclude light dark matter.","marker":"[27]"},{"why":"Supplies the second experimental upper bound on Higgs invisible decay used in the same exclusion.","marker":"[28]"},{"why":"Generates the model files used for the numerical calculation of the relic abundance.","marker":"[31]"},{"why":"Computes the relic abundance including the annihilation and semi-annihilation channels.","marker":"[32]"}],"fun_headline_variants":["Z3-stabilized pNGB dark matter via semi-annihilation","Three complex scalars yield Z3-protected dark matter","Semi-annihilating pNGB dark matter from Z3 symmetry","Pseudo-Goldstone dark matter stabilized by Z3"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes the vacuum in which all three dark scalars get equal background values is the global minimum, but it only checks stationarity and discards a competing vacuum because it breaks the stabilizing Z3 symmetry, without comparing which minimum is deeper.","fun_headline_variants_meta":{"raw":{"variants":["Z3-stabilized pNGB dark matter via semi-annihilation","Three complex scalars yield Z3-protected dark matter","Semi-annihilating pNGB dark matter from Z3 symmetry","Pseudo-Goldstone dark matter stabilized by Z3"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000407,"raw_usage":{"total_tokens":2137,"prompt_tokens":990,"completion_tokens":1147,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":1072}},"tokens_in":606,"tokens_out":1147,"duration_ms":9799,"temperature":1.0,"reasoning_tokens":1072,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:41:42.605288+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the scalar potential at the competing stationary point $v_{s1} \\neq 0$, $v_{s2} = 0$, $v_{s3} = -v_{s1}$ from Appendix A for a benchmark point and compare it with the equal-VEV vacuum; if the Z3-breaking point is deeper, the $a_\\omega$ state is not the ground state and the model's dark-matter conclusion fails.","supporting_citations":[],"review_version":1}