{"id":"ac6b0d5e-45cd-46dd-8764-22924e2eb3bc","arxiv_id":"1208.0009","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Coannihilation, forbidden channels, and asymmetric dark matter can each reconcile a 130 GeV dark matter gamma-ray line with the observed relic density and with continuum gamma-ray limits.","lead":"The paper proposes three mechanisms by which dark matter could produce a bright gamma-ray line near 130 GeV while still having the right cosmic abundance and staying under limits on the broader gamma-ray glow. It is a model-building response to a 2012 Fermi telescope hint that the authors themselves note is unconfirmed and in tension with null line searches.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The reader's equilibrium assumption is the right target, but the paper's own numbers show the coannihilation and forbidden-channel results are robust to it; the sharper soft spot is the asymmetric-DM oscillation window, which is asserted rather than derived.","rationale":"The reader's weakest_assumption (Eq. 3 equilibrium) is a legitimate general worry and correctly identifies the machinery the coannihilation and ADM freeze-out arguments rest on. But applied to this paper it is only partly load-bearing: the coannihilation models in Sec. II use moderate splittings (7-10 GeV against T_f ~ 6 GeV) and electromagnetic-strength conversion couplings, so the equilibrium assumption is plausibly satisfied and the quoted contours are not obviously fragile. The reader's other stated concerns (no shipped code, asserted 1-3% agreement with a numerical solver, unconfirmed line) are real but are correctly weighted as not breaking the conditional claim. My sharper concern is the ADM oscillation washout in Sec. IV, where the sub-4x10^-8 eV splitting and the 3.5% asymmetry are asserted rather than derived, and where the observable line rate depends on that washout. That is a gap in one of three parallel exceptions, not in the central synthesis, and the paper is explicit that the line itself is unconfirmed. So the ACCEPT verdict stands; I would not move it. The concrete test above would settle whether the ADM exception closes, and if it does not, the paper's claim reduces to two working exceptions plus one conditional one, which is a scope reduction rather than a refutation.","tokens_in":55901,"tokens_out":99140,"duration_ms":1917431,"concrete_test":"Solve the coupled Boltzmann equations for the complex scalar chi and its antiparticle including the U(1)_chi-breaking mass term, for m_chi = 130 GeV, g_L = 0, g_R ~ O(1), and splittings spanning 10^-10 to 10^-6 eV. Check whether there is any splitting for which (a) the asymmetry survives freeze-out, (b) oscillations begin after freeze-out, and (c) the late-time chi chi† -> gamma gamma rate reaches 10^-27 cm^3/s without the residual symmetric density exceeding the continuum bound. If no such window exists, the ADM exception as constructed does not close.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader flags Eq. (3) chemical equilibrium as the load-bearing assumption. That is a fair target in general, but the paper's own parameter choices largely defuse it for two of the three exceptions. In Sec. II.A the required splitting is 7-10 GeV with T_f ~ 6 GeV, so exp(-Delta m/T_f) is not extreme and chi_1 f <-> chi_2 f conversion via the same dipole coupling that sets the line rate is fast; in Sec. II.B the text explicitly notes the relic density is set through electromagnetic interactions with 10^-7 << g' << 10^-1, again keeping conversion rates above Hubble. The genuinely under-supported step is in Sec. IV: the claim that chi <-> chi† oscillations wash out the asymmetry after freeze-out requires a U(1)_chi-breaking mass splitting below H(T_f) ~ 4x10^-8 eV, and the paper simply asserts this window exists and that a ~3.5% baryon-asymmetry-sized primordial asymmetry then reproduces Omega_dm. No oscillation dynamics, no Boltzmann equation for the asymmetry, and no estimate of the residual symmetric population from late-time oscillation-induced annihilation are given. Since the entire observable gamma-gamma rate in the ADM scenario depends on that washout having occurred, this is the least secure link in the third exception. Note also that the paper's own Sec. I concedes the 130 GeV line is unconfirmed and in tension with Fermi null searches [10], so all three exceptions are conditional on an observational premise the authors acknowledge.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript addresses the tension between a possible ~130 GeV gamma-ray line signal from the Galactic center (requiring <sigma v>_γγ ~ 10^-27 cm^3/s) and the two standard obstacles for a thermal WIMP interpretation: (i) the relic density requires a total annihilation rate near 3x10^-26 cm^3/s, and (ii) continuum photon constraints bound annihilation to charged SM states at the few x 10^-25 cm^3/s level, while loop-induced γγ is expected to be suppressed relative to tree-level f fbar / WW by (π/α)^2 ~ 10^5. The paper proposes three generic mechanisms that decouple the line rate from the tree-level annihilation rate: coannihilation with a ~O(10 GeV) heavier state that is absent today (illustrated with magnetic dipolar DM and with a coannihilating charged partner), forbidden channels where DM annihilates to slightly heavier charged fermions that are kinematically blocked in the halo but accessible during freeze-out, and asymmetric DM where the relic density is set by a primordial asymmetry later washed out by oscillations. For each, simple Lagrangians are written, relic densities are computed with the standard Griest-Seckel formalism, and parameter regions satisfying both the line rate and Omega h^2 = 0.11 are exhibited.","tokens_in":56957,"tokens_out":100105,"duration_ms":1939422,"significance":"The paper is a useful, clearly organized contribution to the model-building literature on the 130 GeV line. Its value is not a new formalism but a taxonomy: it identifies three structurally distinct ways to reconcile a large loop-induced γγ rate with the relic density and continuum bounds, and demonstrates each with an explicit, renormalizable model. The relic-density calculations use standard machinery (Eqs. 2-5, 20-22) and the one-loop γγ rates are quoted from Bergstrom-Ullio [58] and applied within their stated domain, so the derivations are reproducible from the text. The scenarios are falsifiable in the near term through LHC searches for the near-degenerate charged states, which the authors flag in Sec. V. The main caveat is that the entire program is conditional on an observational premise the authors themselves acknowledge as unconfirmed and in tension with Fermi null searches [10].","major_comments":[{"comment":"The asymmetric DM exception rests on a step that is asserted rather than derived. The text states that χ↔χ† oscillations must begin during or after freeze-out, requiring a U(1)_χ-breaking mass splitting below H(T_f) ~ 4x10^-8 eV, and that if oscillations occur much later the relic density is fixed by an initial asymmetry of O(3.5%) of the baryon asymmetry. No oscillation dynamics, no Boltzmann equation for the asymmetry evolution, and no estimate of the residual symmetric population generated by late-time oscillation-induced annihilation are given. Since the observable γγ rate in this scenario depends on the washout having occurred at the right time, this is the least supported link in the third exception. The authors should either present the oscillation equations and the resulting asymmetry evolution, or explicitly downgrade this to a plausibility argument and state the range of mass splittings and initial asymmetries for which the scenario works.","section":"Sec. IV, Eqs. (31)-(38) and the paragraph after Eq. (38)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The coannihilation relic-density calculation assumes that the individual species densities track their equilibrium ratios r_i, maintained by rapid χ_i f ↔ χ_j f and χ_i ↔ χ_j f fbar conversion. This assumption is load-bearing for both models in Sec. II, since the entire suppression argument is exp(-Δm/T_f) acting on the equilibrium ratio. The paper should state explicitly the condition under which the conversion rates exceed Hubble for the quoted parameters, particularly in Sec. II.A where the required splitting is 7-10 GeV and T_f ~ 6 GeV, and in Sec. II.B where the relic density is set through electromagnetic interactions with 10^-7 << g' << 10^-1. A short quantitative check (or a citation to the standard criterion) would make the results robust rather than assumed.","section":"Sec. II, Eq. (3) and the equilibrium assumption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The forbidden-channel result relies on the resonance at m_φ ≈ 2m_χ to boost γγ while F fbar annihilation occurs at s ≈ 4m_F^2, off the pole. The width Γ_φ entering the propagator is computed from Eq. (25), but the sensitivity of the quoted m_F ~ 130-165 GeV to the resonance treatment is not quantified. Near a narrow resonance the thermal average is sensitive to the precise treatment of the pole and to the velocity distribution; the paper states it uses the full thermal average of Eq. (20), which is appropriate, but should show how the allowed m_F range shifts if the width is varied, e.g. by the opening of additional decay channels above m_φ ~ 300-350 GeV noted in the text. This is a correctness-risk concern about the robustness of the quoted parameter window, not a challenge to the mechanism.","section":"Sec. III, Eq. (30) and Fig. 7","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The estimate <sigma v>_f fbar,WW / <sigma v>_γγ ~ (π/α)^2 ~ 10^5 is presented as generic; it would help to note explicitly that this is an order-of-magnitude expectation for the dimension-six and dimension-seven operators and that the SU(2)_L multiplet arrangement mentioned in the following sentence can change it, so the reader does not take it as a hard bound.","section":"Sec. I, Eq. (1)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The scaling of the required dipole with m_1 in Eq. (9) should be cross-checked against Eq. (8a); the displayed m_1 dependence appears to combine the explicit m_1^2 in the cross section with the fixed target value, and a brief statement of how the scaling was obtained would avoid ambiguity.","section":"Sec. II.A, Eq. (9)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The functions I_n(a,b) are defined only by reference to [58]; since the exact expression is used rather than the m_φ >> m_1,2 limit, it would be helpful to state the domain of a and b over which the numerical implementation was checked against the limit F_+ ≈ 2 - π^2, F_- ≈ 2.","section":"Sec. II.B, Eq. (17) and Eq. (18)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The vertical and horizontal bands indicating the preferred <sigma v>_γγ are helpful; the figure captions should state explicitly which band corresponds to which quantity and whether the error bar is the one quoted from [1], so the figures are self-contained.","section":"Fig. 3 and Fig. 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The numerical prefactors in Eq. (34) should be accompanied by the assumed values of m_F and m_χ used to obtain them, since the quoted rates depend on those choices.","section":"Sec. IV, Eq. (34)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The statement that the models share the feature of new charged states near the DM mass is a useful summary; it would strengthen the paper to note explicitly which of the three scenarios predicts states within LHC reach and which are more model-dependent, since this is the main near-term test.","section":"Sec. V, conclusions","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The citation to [58] for the one-loop γγ cross section is appropriate; given that the exact expression is used, a brief note on which of [58-60] supplies which piece would help the reader reproduce the numbers.","section":"References","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is a well-organized taxonomy of three mechanisms and is likely to be useful and cited. The main substantive gap is the asymmetric-DM oscillation discussion in Sec. IV, which is asserted rather than derived; I would ask for either the oscillation equations or an explicit downgrade to a plausibility argument. The equilibrium assumption in Sec. II deserves a short quantitative check but is probably fine for the quoted parameters. The observational premise (the 130 GeV line) is acknowledged by the authors as unconfirmed and in tension with Fermi null searches, so the paper's value is conditional and should be read as model-building contingent on that signal."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"The short version: this is a well-executed model-building paper, not a discovery paper. The three mechanisms — coannihilation, forbidden channels, asymmetric DM — are all borrowed from the cited literature (Griest-Seckel, Jackson et al., Kaplan-Luty-Zurek). What's new is the explicit 130 GeV application: three concrete Lagrangians, one-loop gamma-gamma rates, and numerical relic-density contours that intersect the line rate with Omega h^2 = 0.11 while staying under the continuum bounds. That intersection is the actual content, and it's done properly. The two constraints are computed independently from the same Lagrangian and then crossed; nothing is tuned to make them meet. The dipolar coannihilation section in particular is tidy — the 7-10 GeV splitting falls out of the relic density rather than being inserted, and the gamma-Z line at 114 GeV is a nice consistency check against the second reported feature. The forbidden-channel section's ratio argument in Eq. (30) is the clearest piece of physics writing in the paper: it shows exactly why the phase-space factor and the pole at m_phi = 2m_chi together beat the loop suppression, and why that only works when m_F sits just above m_chi. Good, honest model building. Where it's soft: the asymmetric DM section. The claim that chi <-> chi-dagger oscillations wash out the asymmetry after freeze-out needs a U(1)_chi-breaking splitting below H(T_f) ~ 4e-8 eV, and the paper asserts that window exists and that a ~3.5% baryon-sized asymmetry then reproduces Omega_dm. No oscillation Boltzmann equation, no estimate of the residual symmetric population from late-time annihilation. Since the whole observable gamma-gamma rate in that scenario depends on the washout having happened, this is the weakest link of the three, and it's the one I'd push on in review. On the reader's Eq. (3) equilibrium worry: it's a fair general target but the paper's own numbers defuse it for the first two exceptions — the splittings and the electromagnetic-strength couplings keep conversion rates above Hubble. I wouldn't make that the headline objection. Minor: no code, and the 1-3% agreement with a numerical solver is stated rather than shown. The 130 GeV premise is flagged as unconfirmed by the authors themselves, which is the right thing to do. Who should read it: anyone building WIMP models for a line signal, and anyone who wants a compact reference for how these three escape hatches actually work at a given mass. It deserves referees.","headline":"A clean, honest synthesis: three known mechanisms applied to the 130 GeV line with relic density and continuum limits intersected, not a new mechanism.","tokens_in":19569,"tokens_out":851,"would_cite":true,"duration_ms":19703,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["95.35.+d","98.70.Rz"],"model":"selfhosted/deepseek-v4.1-flash","headline":"Three loopholes let a thermal dark matter particle produce a 130 GeV gamma-ray line without over-annihilating into ordinary charged particles, keeping both the relic abundance and the continuum photon limits intact.","keywords":["dark matter","gamma-ray lines","thermal relic density","coannihilation","forbidden channels","asymmetric dark matter","WIMP annihilation","Fermi LAT"],"falsifier":"A confirmed continuum gamma-ray excess from the galactic center, halo, or dwarf galaxies at the level implied by a tree-level fermion channel would rule out the coannihilation and forbidden-channel realizations, and a Fermi LAT line search with more data that excludes $⟨σv⟩_{γγ} ≈ 10^{-27}$ cm$^3$/s would remove the target all three models are built to hit.","tokens_in":20035,"feed_emoji":"🌌","tokens_out":2573,"duration_ms":42945,"temperature":0.7,"pith_summary":"A faint feature near 130 GeV in Fermi Large Area Telescope data has been read by several groups as a dark matter annihilation line into two photons, requiring a cross section of about $10^{-27}$ cm$^3$/s. The trouble is that dark matter has no electric charge, so a photon coupling must be generated through a loop of charged particles, and the same charged particles give a tree-level annihilation channel into fermions or $W$ pairs that is larger by roughly $(π/α)^2 ≈ 10^{5}$. That would both deplete the relic density far below what is observed and produce a continuum gamma-ray flux already excluded by Fermi and dwarf-galaxy searches. The paper's claim is that three specific mechanisms break this chain: coannihilation with a heavier near-degenerate state that has left the thermal bath by today, annihilation into charged states slightly heavier than the dark matter so that it is kinematically shut off in the halo but open during freeze-out, and a primordial asymmetry that quenches annihilation in the early Universe and is later erased by particle-antiparticle oscillations. For each mechanism the authors build an explicit model, compute the relic density, and show the required parameters are consistent with continuum bounds. The value of the work is that it turns a general objection — a large loop-induced photon coupling is incompatible with a thermal WIMP — into a list of three concrete ways out, each with its own distinctive prediction for the mass spectrum.","feed_headline":"Three loopholes let dark matter shine in gamma rays","feed_subtitle":"Coannihilation, forbidden channels and asymmetry each explain a 130 GeV line while keeping the relic density right.","key_machinery":"The load-bearing machinery is the separation between the annihilation channel that sets the relic abundance and the channel that produces the line, achieved by a Boltzmann or phase-space suppression that is active in one epoch and not the other. Quantitatively this is the effective thermally-averaged cross section $⟨σ_{eff}v⟩ = Σ_{i,j} r_i r_j ⟨σ_{ij}v⟩$ with equilibrium ratios $r_i ∝ g_i (1+Δ_i)^{3/2} e^{-xΔ_i}$, entering $Ω_{dm}h^2 = (1.07×10^9 GeV^{-1}/(g_*^{1/2} m_{Pl})) [∫_{x_f}^{∞} x^{-2} ⟨σ_{eff}v⟩ dx]^{-1}$. Coannihilation suppresses the fermion channel by $e^{-Δm/T_f}$; the forbidden channel suppresses it by the final-state phase space $v_2^2$ with a minimum velocity $2(1-m_χ^2/m_F^2)^{1/2}$; the asymmetric case suppresses it by the chemical potential plus a $p$-wave or chirality factor. In each case the loop amplitude for $χχ → γγ$ — dimension six for scalars, dimension seven for fermions, or the dipole-mediated $t$-channel — is untouched by the suppression.","core_discovery":"The paper argues that the apparent impossibility of a thermal WIMP producing a $10^{-27}$ cm$^3$/s two-photon line is not a theorem but a consequence of assuming that the same charged particles mediating the loop are kinematically accessible and thermally populated in both the early Universe and the halo today. Relaxing that assumption in three ways restores consistency. In coannihilation, freeze-out is driven by $χ_1χ_2 → f̄f$ with a mass splitting of order 10 GeV, so the fermion channel is Boltzmann-suppressed at freeze-out and absent today once $χ_2$ has decayed away; the authors realize this with magnetic dipolar dark matter, where a dipole of about $3.6×10^{-3}μ_N$ gives the line and a 7–10 GeV splitting gives $Ω h^2 = 0.11$, and with a charged coannihilating partner generating the photon coupling at dimension seven. In the forbidden case, dark matter annihilates into a charged fermion $F$ heavier than itself; the early-Universe velocity tail opens the channel while the halo velocity does not, and the authors show a pseudoscalar-mediated model needs $m_F$ near 130–165 GeV, with an $s$-channel resonance at $m_φ ≈ 2m_χ$ reducing the required coupling. In the asymmetric case, a primordial $χ$–$χ^†$ asymmetry quenches $χχ^† → f̄f$ after freeze-out, and oscillations erase the asymmetry so that annihilation can proceed today; a chiral scalar coupling makes the fermion channel $p$-wave or chirality-suppressed, leaving the loop-induced $γγ$ rate of $10^{-27}$ cm$^3$/s compatible with continuum limits for order-one couplings.","pith_inferences":["The same logic inverts into a search strategy: any DM model that explains a photon line through loops while remaining thermal must hide its tree-level channel in one of these three ways, so the three exceptions are close to a classification of viable thermal line models rather than a menu of examples.","Because all three mechanisms require new charged states near the DM mass, the line signal and the absence of a continuum excess together act as a mass-spectrum measurement, not merely a coupling measurement.","If the 130 GeV feature proves instrumental, the framework still stands as a template for how a future line detection could be reconciled with thermal freeze-out, so the value of the paper does not reduce to the fate of that one feature."],"forward_implications":["A dark matter interpretation of the 130 GeV line survives without violating relic-density or continuum-photon bounds, provided the spectrum contains a near-degenerate or slightly heavier charged state within roughly 10–35 GeV of the dark matter mass.","The three scenarios predict new charged particles in the 130–165 GeV range, accessible to LHC searches and able to shift the Higgs diphoton rate, giving a collider handle distinct from the line itself.","The coannihilation and forbidden-channel models predict no annihilation to fermions in the halo today, so a confirmed continuum excess accompanying the line would disfavour them in favour of the asymmetric scenario or of non-thermal production.","The asymmetric model requires a $U(1)_χ$-breaking mass splitting below about $4×10^{-8}$ eV so that oscillations begin at or after freeze-out, tying the line signal to a very small explicit symmetry-breaking scale."],"supporting_citations":[{"why":"Supplies the 130 GeV line signal and the target cross section $⟨σv⟩_{γγ} ≈ 10^{-27}$ cm$^3$/s that the models are built to reproduce.","marker":"[1]"},{"why":"Fermi LAT line and continuum searches that set the upper bounds the models must evade in the halo today.","marker":"[10]"},{"why":"Griest–Seckel formalism for coannihilation and forbidden-channel relic density, the computational backbone of Eqs. (3)–(5) and (20)–(22).","marker":"[53]"},{"why":"Prior dipolar dark matter treatment with small mass splitting, the baseline the coannihilation model departs from.","marker":"[36]"},{"why":"Earlier use of the forbidden-channel mechanism with the top quark to generate $γZ$ and $γh$ lines, which this paper adapts to $γγ$.","marker":"[8]"},{"why":"Kaplan–Luty–Zurek asymmetric dark matter, the framework underlying Sec. IV.","marker":"[61]"},{"why":"Oscillation-induced washout of the asymmetry and its effect on relic density, which the asymmetric model relies on to allow annihilation today.","marker":"[70]"},{"why":"Loop amplitude for $χχ → γγ$ used in Eq. (17) for the coannihilation-with-charged-partners model.","marker":"[58]"}],"fun_headline_variants":["Three exceptions keep thermal dark matter compatible with the line","Coannihilation, forbidden channels and asymmetry can rescue the line","A 130 GeV photon line does not rule out thermal dark matter","Thermal dark matter can evade continuum limits in three ways","Forbidden channels and asymmetry can open the photon-line path"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The relic-density and coannihilation calculations assume that all the nearly degenerate species stay in chemical equilibrium with each other through rapid conversion processes, so that their number densities track the Boltzmann-weighted ratios $r_i$ of Eq. (3); if those conversion rates fall below the Hubble rate, the effective cross section used to derive every quoted mass splitting is not the relevant quantity.","fun_headline_variants_meta":{"raw":{"variants":["Three exceptions keep thermal dark matter compatible with the line","Coannihilation, forbidden channels and asymmetry can rescue the line","A 130 GeV photon line does not rule out thermal dark matter","Thermal dark matter can evade continuum limits in three ways","Forbidden channels and asymmetry can open the photon-line path"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2233,"prompt_tokens":1093,"completion_tokens":1140,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":1070}},"tokens_in":1093,"tokens_out":1140,"duration_ms":19033,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":1070,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-10T13:56:17.517039+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"A confirmed continuum gamma-ray excess from the galactic center, halo, or dwarf galaxies at the level implied by a tree-level fermion channel would rule out the coannihilation and forbidden-channel realizations, and a Fermi LAT line search with more data that excludes $⟨σv⟩_{γγ} ≈ 10^{-27}$ cm$^3$/s would remove the target all three models are built to hit.","supporting_citations":[{"cited_title":"A Tentative Gamma-Ray Line from Dark Matter Annihilation at the Fermi Large Area Telescope","cited_arxiv_id":"1204.2797","evidence_quote":"Supplies the 130 GeV line signal and the target cross section $⟨σv⟩_{γγ} ≈ 10^{-27}$ cm$^3$/s that the models are built to reproduce."},{"cited_title":"Fermi LAT Search for Dark Matter in Gamma-ray Lines and the Inclusive Photon Spectrum","cited_arxiv_id":"1205.2739","evidence_quote":"Fermi LAT line and continuum searches that set the upper bounds the models must evade in the halo today."},{"cited_title":"Phys.Rev. D43, 3191","cited_arxiv_id":null,"evidence_quote":"Griest–Seckel formalism for coannihilation and forbidden-channel relic density, the computational backbone of Eqs. (3)–(5) and (20)–(22)."},{"cited_title":"How Dark Are Majorana WIMPs? Signals from MiDM and Rayleigh Dark Matter","cited_arxiv_id":"1206.2910","evidence_quote":"Prior dipolar dark matter treatment with small mass splitting, the baseline the coannihilation model departs from."},{"cited_title":"Higgs in Space!","cited_arxiv_id":"0912.0004","evidence_quote":"Earlier use of the forbidden-channel mechanism with the top quark to generate $γZ$ and $γh$ lines, which this paper adapts to $γγ$."},{"cited_title":"Asymmetric Dark Matter","cited_arxiv_id":"0901.4117","evidence_quote":"Kaplan–Luty–Zurek asymmetric dark matter, the framework underlying Sec. IV."},{"cited_title":"Consequences of DM/antiDM Oscillations for Asymmetric WIMP Dark Matter","cited_arxiv_id":"1110.3809","evidence_quote":"Oscillation-induced washout of the asymmetry and its effect on relic density, which the asymmetric model relies on to allow annihilation today."},{"cited_title":"Full One-loop Calculation of Neutralino Annihilation into Two Photons","cited_arxiv_id":"hep-ph/9706232","evidence_quote":"Loop amplitude for $χχ → γγ$ used in Eq. (17) for the coannihilation-with-charged-partners model."}],"review_version":1}