{"id":"c794d4b2-2be5-40b6-936a-45a13d054f01","arxiv_id":"2412.00523","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"The CP-odd scalar A in the flavor-aligned CP-violating 2HDM can explain the diphoton excesses at 95 and 152 GeV only in tuned regions that survive the electron EDM bound and predict measurable neutron and proton EDMs.","lead":"A CP-violating two-Higgs-doublet model can explain the 95 GeV and 152 GeV diphoton excesses with decays of the CP-odd scalar A, while the same complex coupling generates electric dipole moments. The proposal links two collider hints to future neutron and proton EDM experiments as decisive tests.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 95 GeV electron-EDM consistency rests on the tuned and unmotivated choice aD = aE = 0 in Sec. V A; a non-negligible aE at the level of aU would restore a Barr-Zee contribution to d_e that risks violating Eq. (33), so the claimed viable region is a fine-tuned slice.","rationale":"I read the paper as claiming the existence of flavor-aligned 2HDM parameter space in which A ≈ h3 explains the 95 and 152 GeV diphoton excesses while evading the electron EDM bound of Eq. (33), with correlated future neutron and proton EDM sensitivity. The scalar-sector machinery (Higgs-basis formalism, Eq. (32) for hk → γγ, Eqs. (34)–(36) for the neutron EDM) is internally consistent; the cross-checks with ScannerS and HiggsTools, the vacuum-stability and unitarity checks, and the use of the public code of Ref. [90] are real independent support. I considered two alternative concerns and found them secondary. (i) Does the 95 GeV benchmark quantitatively reach μ_γγ = 0.24? Using the printed Eq. (39) and Table II, an order-of-magnitude reconstruction gives σ × BR ≈ 0.02 pb, close to the 0.024 pb needed for the observed signal, so the blue band of Fig. 2 is plausible; this is a reproducibility limitation, not a demonstrated failure. (ii) The 152 GeV scenario's BR(h3 → W±H∓) = 0.95 requires the W to be off-shell for m_h3 = 152 GeV and m_H± = 130 GeV; the paper explicitly acknowledges this ('with, say, the W boson off-shell'), so there is no internal inconsistency. The genuine load-bearing weakness is the one the reader identified, and it is disclosed by the paper itself: electron-EDM evasion is bought by setting aD = aE = 0 (Sec. V A, Table III) or by keeping |aF| ≤ 0.01 with a small phase (Sec. V B). Because aU, aD, aE are free, independent parameters that renormalize (Eq. (28)), the hierarchy aE ≪ aU has no symmetry or UV justification in the paper. The decisive check is a single re-computation of d_e as a function of aE/aU and their relative phase. If the check shows the bound is violated for aE ~ aU/10, the CONDITIONAL verdict stands and the authors should be asked to quantify and motivate the tuning; if the check refutes the concern, the verdict should be upgraded. My read therefore leaves the reader's CONDITIONAL verdict unchanged.","tokens_in":22373,"tokens_out":37759,"duration_ms":378766,"concrete_test":"Recompute d_e with the public code of Ref. [90] (or an independent Barr-Zee implementation) at the Sec. V A benchmark (m_h3 = 95 GeV, m_H± = 130 GeV, θ12 = 0.25, θ13 = 0.01, Z3 = −0.2, Re Z7 = 0.1, Im Z7 = 0.4, Re aU = −0.01, Arg aU = −0.03) for aD = aE = r |aU| e^{iδ} with r ∈ {0, 10⁻³, 10⁻², 10⁻¹, 1} and δ ∈ {0, −0.03, π/2}. If |d_e| exceeds 1.1 × 10⁻³⁰ e cm for any r ≥ 0.1, the electron-EDM consistency of the 95 GeV region rests on tuning aE and aD at least one order of magnitude below aU without UV motivation. If instead |d_e| stays below the bound for all r up to 1 and all δ, the tuning objection is refuted and the verdict should be upgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that A ≈ h3 can explain the 95 (and 152) GeV diphoton excesses while remaining consistent with the electron EDM bound of Eq. (33). The mechanism that makes this consistency possible is stated explicitly in Sec. V A: 'let us consider the dependence on aU and Im Z7 while setting aD = aE = 0, which strongly suppresses an effect in the very constraining electron EDM.' This is the load-bearing step. In the flavor-aligned 2HDM the parameters aU, aD, aE of Eq. (28) are independent complex numbers; no symmetry forces the hierarchy |aD|, |aE| ≪ |aU| ≈ 0.01, and the alignment-stability literature cited (Refs. [84,85]) bounds departures from exact alignment, not the aE ≪ aU hierarchy. Since d_e gets two-loop contributions scaling as Im Z7 × Im(aE) and Im(aE aU*), a generic aE of order |aU| = 0.01 with misaligned phase would push d_e toward or above 1.1 × 10⁻³⁰ e cm, cutting into most of the blue region of Fig. 2; the benchmark is only safe because Im aE ≈ −3 × 10⁻⁴ is suppressed by the chosen phase and by aE = 0. The same tuning appears in the 152 GeV case (Sec. V B), where the electron EDM is the dominant constraint (Fig. 3) and forces |aF| ≤ ~0.01 with Arg aF = −0.01. The advertised consistency with Eq. (33) therefore holds only in a specially chosen, radiatively unprotected slice of parameter space; the A → γγ versus EDM correlation that is the paper's headline is not generic to the CP-violating flavor-aligned 2HDM. No internal inconsistency in the loop calculations was found; the concern is the robustness of the claimed viable region.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper examines the CP-violating flavor-aligned two-Higgs-doublet model in the Higgs basis, with the observed 125 GeV Higgs in the alignment limit. It observes that a complex Z7 parameter generates an H+H-h3 coupling that can make the decay h3 ≃ A → γγ sizable, and that the same CP-violating source induces electron, neutron, and proton EDMs. For two benchmark scenarios, m_h3 = 95 GeV and m_h3 = 152 GeV, the paper claims that the LHC diphoton excesses can be explained while respecting the current electron EDM bound, and that future neutron and proton EDM experiments will probe the relevant parameter space. It also comments on a possible h2 ≃ H interpretation of the LEP 98 GeV excess.","tokens_in":22816,"tokens_out":19464,"duration_ms":180160,"significance":"The paper's main contribution is a concrete, falsifiable link between the observed diphoton excesses and future EDM searches within a well-defined BSM framework. It provides a basis-invariant treatment of the CP-violating scalar sector and explicit branching-ratio tables for the benchmarks, and it checks several constraints with public tools (ScannerS, HiggsTools, MadGraph). The central mechanism—using a complex Z7 to generate both a large A→γγ rate and correlated EDMs—is interesting and potentially important, and the paper identifies parameter regions that future neutron and proton EDM experiments can decisively test. The caveat is that the quantitative regions shown rely on specially chosen alignment parameters and phases, as detailed below; if the result survives a broader scan, it would constitute a significant step in interpreting the 95 and 152 GeV excesses.","major_comments":[{"comment":"The viability of the 95 GeV explanation rests on the choice, stated after Eq. (39), to set aD = aE = 0, which 'strongly suppresses an effect in the very constraining electron EDM.' In the flavor-aligned 2HDM of Eq. (28), aU, aD, and aE are independent complex numbers; no symmetry in the model forces |aD|, |aE| to be much smaller than |aU| ~ 0.01. The two-loop Barr-Zee contributions to de scale as Im Z7 × Im(aE) and Im(aE aU*), so a generic aE of order aU with a non-negligible phase would violate the bound of Eq. (33) over most of the blue region in Fig. 2. The benchmark in Table III is only safe because aE = 0 and Arg aU = -0.03 are chosen. Since the abstract's headline claim is consistency with the electron EDM bound, the paper should either justify this hierarchy with a UV symmetry or extend the scan to non-zero aD and aE and quantify the surviving parameter space. As it stands, the advertised correlation is demonstrated only on a fine-tuned slice of the model's parameter space.","section":"Sec. V A, Table III"},{"comment":"In the 152 GeV analysis, the authors fix aU = aD = aE = aF and choose Arg aF = -0.01 (Table V). The resulting electron-EDM constraint (orange region in Fig. 3) is satisfied because this small phase suppresses Im aF and thus the contributions to de. For a generic phase of order one, de would receive unsuppressed two-loop contributions proportional to Im Z7 × Im(aF) and Im(aF × aU*), and the allowed region in Fig. 3 would shrink substantially. The paper does not scan over the phase of aF or show that the correlation is phase-independent; it therefore establishes the A→γγ/EDM correlation for the 152 GeV excess only on a second, independently tuned slice. A scan or an analytic argument demonstrating insensitivity to the phase is needed to support the general claim.","section":"Sec. V B, Table V, Fig. 3"}],"minor_comments":[{"comment":"The chain of approximations 'σGF(pp→h3) ≈ [1.5/(1+s12^2)] σGF(pp→h95) ≈ |aU|^2 100 pb' is not dimensionally consistent as typeset, since the first expression does not contain |aU|^2. The intended formula is likely σGF(pp→h3) ≈ [1.5/(1+s12^2)] |aU|^2 σGF(pp→h95) with σGF(pp→h95) ≈ 100 pb, and the text should be corrected.","section":"Sec. III, Eq. (39)"},{"comment":"The notation '|de| ≤ (1.3 ± 2.0stat ± 0.6sys) × 10^-30 e cm' is not a valid statement of a bound, because an upper limit does not carry a central value and uncertainties. Please quote the measured value and the 90% CL limit separately, or clarify the exact statistical treatment used in the figures.","section":"Sec. IV, Eq. (33)"},{"comment":"The ScannerS cross-check is performed only for λ6 = λ7 = 0 and for the type-I Yukawa sector; it does not exercise the CP-violating Z7 couplings or the flavor-aligned Yukawa structure that drive the main results. The authors should state more precisely what the cross-check validates, or provide an additional check of the alignment-limit H+H-hk couplings used in the analysis.","section":"Sec. III, after Eq. (32)"},{"comment":"The electron EDM calculation is taken from the code of Ref. [90] rather than derived or displayed. Since the electron EDM bound is a central observable in this paper, including the analytic expression in the text or an appendix would improve reproducibility and allow the reader to see the parameter dependence explicitly.","section":"Sec. IV"},{"comment":"The benchmark point with Re aU = -0.01 and Im Z7 = 0.4 yields a gluon-fusion production cross section that, combined with BR(h3→γγ) = 0.037, falls far below the signal required for the 95 GeV excess [cf. Eq. (37)]. It would clarify the presentation to state explicitly that this benchmark is used only for the branching-ratio table and does not lie in the preferred blue region of Fig. 2, or to choose a benchmark that is representative of the viable region.","section":"Sec. V A, Tables II and III"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope, and the central formalism is sound. The main concern is the naturalness of the parameter choices that evade the electron EDM; I recommend major revision rather than rejection, because the issue is addressable with an extended scan and a more careful statement of the claims. The authors should also fix the notation errors in Eqs. (33) and (39)."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is the correlation itself: a single CP-violating parameter Z7 in the flavor-aligned 2HDM produces both an enhanced A→γγ rate and electron, neutron, and proton EDMs, and the paper applies this to the 95 and 152 GeV diphoton excesses. That is a genuinely useful observation, and the diphoton/EDM machinery is standard and mostly cross-checked with public codes (ScannerS, HiggsTools, the Altmannshofer-Gori-Hamer-Patel EDM code). For a phenomenological paper in this crowded subfield, the analysis is careful and the parameter counting is transparent.\n\nThe main soft spot is the one the stress-test note identifies: the 95 GeV viability depends on setting aD = aE = 0 (Sec. V A, Table III), and the 152 GeV case also needs a universal aF small enough to evade the electron EDM. Nothing in the flavor-aligned construction forces that hierarchy; it is a fine-tuned slice, and the paper offers no symmetry or UV argument for it. That does not make the correlation wrong, but it does mean the headline claim—\"regions of parameter space that can be tested with future neutron and proton EDM measurements\"—is less robust than the wording suggests. The electron EDM bound is doing the real work, and shifting aE by an order of magnitude likely wipes out most of Fig. 2. I also could not fully reproduce the figures from the text alone; the scans are not documented in enough detail.\n\nOther concerns are minor: the electron EDM is taken from an external code rather than re-derived here (fine for a phenomenology paper), the hadronic uncertainties in the neutron/proton EDM projections are noted but not propagated, and the whole exercise assumes the excesses are real. I do not see any internal inconsistency in the loop calculations, and the paper is honest about the LEP/98 GeV connection being a secondary possibility.\n\nBottom line: this is a competent, worth-reading contribution to the 95/152 GeV excess literature, and the correlation it highlights is plausible. It should go to peer review, but the referee should push for reproducible scans and a straight discussion of the aD=aE=0 tuning before endorsing the viable regions.","headline":"Plausible new correlation between A→γγ and EDMs in the 2HDM, but the advertised viable regions rely on an unmotivated tuning of the flavor-alignment parameters.","tokens_in":23406,"tokens_out":1057,"would_cite":true,"duration_ms":13148,"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":"A single complex parameter of the two-Higgs-doublet model could produce both observed diphoton excesses, tying a 95 GeV or 152 GeV $A\\to\\gamma\\gamma$ signal to electron, neutron, and proton electric dipole moments.","keywords":["two-Higgs-doublet model","flavor alignment","CP violation","diphoton excess","electric dipole moment","Higgs alignment limit","charged Higgs loop","light Higgs boson"],"falsifier":"Measure the neutron EDM at about $10^{-27}\\,e\\,\\mathrm{cm}$ and the proton EDM at about $10^{-29}\\,e\\,\\mathrm{cm}$: the paper predicts that the parameter space explaining the 95 GeV excess (and, with $a_E=0$, the 152 GeV excess) gives EDMs at or above these levels, so null results would rule out the mechanism. A complementary test is a future electron EDM measurement near $10^{-31}\\,e\\,\\mathrm{cm}$, which with nonzero $a_E$ would exclude the 152 GeV region shown in Fig. 3.","tokens_in":22139,"feed_emoji":"⚛️","tokens_out":7911,"duration_ms":69943,"temperature":0.7,"pith_summary":"This paper proposes that the diphoton excesses seen at 95 GeV and 152 GeV at the LHC could both be produced by decays of the mostly CP-odd scalar $h_3\\simeq A$ in the CP-violating flavor-aligned two-Higgs-doublet model. The same complex scalar-potential parameter $\\bar Z_7$ that gives $A$ a sizable branching ratio to two photons, through a charged-Higgs loop, also generates electric dipole moments for the electron, neutron, and proton. The authors show that the 95 GeV and 152 GeV excesses can be explained while respecting the current electron EDM bound, and that the surviving parameter regions fall within the reach of planned neutron and proton EDM experiments. If the 95 GeV excess is really $A$, then the mostly CP-even partner $h_2\\simeq H$ at about 98 GeV can simultaneously explain the LEP Higgs-strahlung excess in $b\\bar b$.","feed_headline":"One parameter could explain both diphoton excesses","feed_subtitle":"The same CP-violating coupling that boosts A→γγ predicts neutron and proton EDMs within reach of next experiments.","key_machinery":"The load-bearing object is the basis-invariant complex parameter $\\bar Z_7$ in the scalar potential of the general two-Higgs-doublet model. In the Higgs-alignment limit the charged-Higgs coupling $H^+H^-h_3$ is $v\\,\\mathrm{Im}\\,\\bar Z_7$, so a nonzero imaginary part produces $A\\to\\gamma\\gamma$ through a charged-Higgs loop without suppression from small scalar mixing; the same $\\bar Z_7$ phase generates CP-violating Yukawa phases that induce fermion EDMs through Barr-Zee two-loop diagrams and the three-gluon Weinberg operator. The flavor-alignment parameters $a_U,a_D,a_E$ control the fermion-loop contributions and, crucially, the electron EDM: setting $a_D=a_E=0$ (95 GeV benchmark) or taking a small universal $a_F$ (152 GeV benchmark) suppresses the electron EDM while keeping the top-quark coupling needed for gluon-fusion production.","core_discovery":"The central claim is that a single CP-violating interaction, the $H^+H^-A$ coupling controlled by $\\mathrm{Im}\\,\\bar Z_7$, can give the mostly CP-odd state $A$ a large enough branching ratio to photons to account for the observed 95 GeV and 152 GeV diphoton excesses, while the same interaction feeds through two-loop Barr-Zee and Weinberg-operator diagrams into electron, neutron, and proton EDMs. Under the Higgs alignment limit and with small flavor-alignment parameters $a_F$, the electron EDM is evaded at the current bound for a benchmark with $a_D=a_E=0$ at 95 GeV and for small universal $a_F$ at 152 GeV, where production is dominated by Drell-Yan. The paper thereby correlates a collider anomaly with low-energy CP-violating observables in a basis-invariant way: the same $\\bar Z_7$ that sets the diphoton rate also sets the EDM predictions.","pith_inferences":["If both excesses are confirmed and this mechanism is right, the scalar spectrum is tightly predicted: $m_A$ at 95 or 152 GeV, $m_H$ near 98 GeV for the first case, and $m_{H^\\pm}$ near 130 GeV, alongside a SM-like 125 GeV Higgs.","The choice $a_D=a_E=0$ suppresses the electron EDM by fiat; a symmetry that enforced this alignment would make the 95 GeV explanation more natural, and without it the electron EDM bound is a serious threat to the model.","The same correlation should hold in other extensions with a light CP-odd scalar: any large $A\\to\\gamma\\gamma$ rate obtained through a charged-scalar loop with CP violation will generically predict nucleon EDMs near current limits.","Future LHC data on $\\gamma\\gamma+X$ and on $H^\\pm\\to cb$ can discriminate this scenario from a CP-conserving interpretation, because here the $A\\to\\gamma\\gamma$ rate and the EDM predictions are locked together."],"forward_implications":["The 95 GeV diphoton excess can be identified with $A$, while $H$ at about 98 GeV accounts for the LEP excess in $e^+e^-\\to ZH$ with $H\\to b\\bar b$.","The 152 GeV diphoton excess can be reproduced only when the product $\\mathrm{Im}\\,\\bar Z_7 \\times |a_F|$ is small enough to evade the electron EDM, forcing production via Drell-Yan rather than gluon fusion.","Future neutron and proton EDM experiments at projected sensitivities should cover most of the parameter space that explains the 95 GeV excess.","For the 152 GeV benchmark the state $A$ has a large branching ratio to $W^\\pm H^\\mp$, so searches for a light charged Higgs boson provide a direct test of the scenario."],"supporting_citations":[{"why":"Establishes the basis-invariant Higgs-basis formalism and the invariant $\\bar Z_7$ on which the argument rests.","marker":"[54]"},{"why":"Defines the flavor-aligned two-Higgs-doublet model with complex alignment parameters $a_F$ used in the benchmark scans.","marker":"[64]"},{"why":"Provides the electron EDM bound that constrains the CP-violating parameter space.","marker":"[27]"},{"why":"Supplies the public electron-EDM calculation code the paper uses for the electron EDM constraint.","marker":"[90]"},{"why":"Reports the ATLAS search whose 95 GeV diphoton excess is a target of the paper.","marker":"[35]"},{"why":"Reports the CMS search corroborating the 95 GeV diphoton excess used in the signal-strength combination.","marker":"[107]"},{"why":"Provides the combined 95 GeV signal strength that defines the target region the model must reproduce.","marker":"[108]"},{"why":"Gives the 152 GeV associated-diphoton analysis whose preferred branching ratio is adopted.","marker":"[53]"},{"why":"Provides the projected proton EDM sensitivity that makes the scenario testable.","marker":"[91]"},{"why":"Introduces the Barr-Zee two-loop diagrams through which $\\bar Z_7$ generates fermion EDMs.","marker":"[92]"}],"fun_headline_variants":["One CP-violating coupling links diphoton excesses to EDMs","Same coupling drives 95 & 152 GeV diphoton bumps and EDMs","Diphoton excesses and EDMs tied to single 2HDM parameter","CP violation unites diphoton bumps with neutron and proton EDMs","A single Im Z7 explains both diphoton excesses and EDMs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The explanation of the 95 GeV excess depends on the ad hoc choice to set the down-type and lepton flavor-alignment couplings to zero, which removes the electron electric dipole moment while keeping a small top-quark coupling for production; if those couplings are instead comparable to the top coupling, the current electron EDM bound excludes most of the favored region.","fun_headline_variants_meta":{"raw":{"variants":["One CP-violating coupling links diphoton excesses to EDMs","Same coupling drives 95 & 152 GeV diphoton bumps and EDMs","Diphoton excesses and EDMs tied to single 2HDM parameter","CP violation unites diphoton bumps with neutron and proton EDMs","A single Im Z7 explains both diphoton excesses and EDMs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000268,"raw_usage":{"total_tokens":1759,"prompt_tokens":1224,"completion_tokens":535,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":840,"completion_tokens_details":{"reasoning_tokens":434}},"tokens_in":840,"tokens_out":535,"duration_ms":4691,"temperature":1.0,"reasoning_tokens":434,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T05:18:24.271751+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the neutron EDM at about $10^{-27}\\,e\\,\\mathrm{cm}$ and the proton EDM at about $10^{-29}\\,e\\,\\mathrm{cm}$: the paper predicts that the parameter space explaining the 95 GeV excess (and, with $a_E=0$, the 152 GeV excess) gives EDMs at or above these levels, so null results would rule out the mechanism. A complementary test is a future electron EDM measurement near $10^{-31}\\,e\\,\\mathrm{cm}$, which with nonzero $a_E$ would exclude the 152 GeV region shown in Fig. 3.","supporting_citations":[],"review_version":1}