{"id":"2798aa81-49a8-4291-b9d8-8b7468f9f1fe","arxiv_id":"1303.3879","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"Combining Moriond 2013 Higgs data, the 126 GeV boson's couplings match the Standard Model in sign and are linear in particle mass to within a few percent, with the symmetry-breaking scale fixed at 244 GeV to within 10%.","lead":"A global fit to the March 2013 LHC and Tevatron Higgs data finds that the new boson's couplings to fermions and to W/Z bosons have the same relative sign, as the Standard Model requires, and that they scale with particle mass linearly to within a few percent. It is a quantitative check of the Higgs mechanism itself, useful to anyone tracking how tightly the 126 GeV particle is pinned to the Standard Model.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The reconstruction discrepancy the reader flags is roughly 0.25σ in a single channel, far too small to move a Δχ²≈9 sign gap; the genuine load-bearing caveat is the assumed SM loop content for c_γ, which the paper itself tests in Fig. 4.","rationale":"The reader correctly senses that the sign result is conditional, and correctly notes that the authors' CMS γγ reconstruction differs from the official combination. But the proposed mechanism — that a small shift in that one channel's central value could move a Δχ² gap of about 9 — does not survive a quantitative check. The discriminating power comes from the loop interference structure: with c<0 the top-loop contribution flips sign relative to the W loop, predicting a γγ rate roughly twice the SM value, which is excluded by both experiments independently of whether μ_γγ is 0.72 or 0.78. The paper's own Fig. 4 right panel identifies the actual condition: marginalizing over c_γ restores the sign symmetry, so the sign claim is conditional on the no-BSM-loop assumption. That is an acknowledged limitation, stated in Section 3, not a hidden one. The reader's verdict of CONDITIONAL is defensible on other grounds (no released likelihood code, digitized inputs not provided, the loop-content assumption), and I would not move it. But the specific weak point named — reconstruction precision of the CMS γγ channel — is not load-bearing at the stated magnitude, and the reader's 'sub-percent-level shift' framing misstates both the size of the discrepancy and its leverage. I therefore record partial agreement: the conditionality is real, the proposed lever is not the right one, and the verdict stands unchanged. The concrete test above would settle it by substituting the official numbers and checking whether the c<0 branch's predicted μ_γγ remains far above the data.","tokens_in":13575,"tokens_out":4108,"duration_ms":103170,"concrete_test":"Replace the reconstructed CMS γγ inputs with the official combination (μ = 0.78 +0.28/−0.26, and the official 7 TeV value 1.69 +0.65/−0.59), rerun the (a,c) fit, and report two numbers: the Δχ² between the c>0 and c<0 minima, and the predicted μ_γγ at the c<0 minimum. If the c<0 minimum predicts μ_γγ ≳ 2 and Δχ² stays ≳ 6, the sign claim is robust to the reconstruction and the reader's concern does not land. If Δχ² falls below ~4, the concern lands and the 'first time' language should be softened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's mechanism is directionally sensible but quantitatively too weak to carry the weight assigned to it. The sign discrimination does not rest on the precise central value of the CMS γγ likelihood; it rests on the fact that the c<0 branch predicts a γγ rate far above the Standard Model. In the c<0 branch the top-loop term flips sign and adds constructively to the (dominant, negative) W-loop amplitude, so with a≈1 the predicted μ_γγ is roughly a factor of two or more above unity. Both ATLAS (1.65) and CMS (0.78) sit at or below the SM value, so the c<0 branch is excluded at many sigma by the qualitative fact of a non-enhanced γγ rate, not by the 0.06 difference between the authors' reconstructed CMS combination (0.72) and the official one (0.78). A 0.06 shift in μ with ±0.25 errors is ~0.25σ and cannot move a Δχ² of 9. The reader's own framing ('sub-percent-level shift') understates the size of the discrepancy while overstating its consequence. The real load-bearing assumption is the one the paper states explicitly in Section 3 and displays in the right panel of Fig. 4: the sign preference is obtained assuming no non-Standard-Model particles contribute to c_γ (and c_g). When c_γ is marginalized, the paper itself reports that 'the symmetry between the solutions with c>0 and <0 is restored.' That is the condition under which the 'first time' sign claim holds, and the authors acknowledge it. So the concern as posed does not land; the conditionality the reader senses is real but is already quantified by the authors, and the reconstruction discrepancy is not the lever.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"Ellis and You perform a global fit to the Higgs boson couplings using the ATLAS, CMS and Tevatron results presented at the Moriond 2013 meetings. Working in the standard effective-Lagrangian framework of Eq. (3) with a common boson rescaling a and a universal fermion rescaling c, and treating the loop-induced couplings c_gamma and c_g as free parameters where appropriate, they find a global minimum at a = 1.03 +/- 0.06 and c = 0.84 +/- 0.15 (Eq. 5), and report that for the first time the data prefer c > 0 over c < 0 at Delta chi-squared of about 9, driven largely by the new CMS photon-photon result. They also fit the two-parameter (M, epsilon) parameterization of Eq. (1), which interpolates between the Standard Model and a power-law mass dependence, obtaining M = 244 +20/-10 GeV and epsilon = -0.022 +0.042/-0.021 (Eq. 7), with a chi-squared price of 0.12 for the Standard Model point. Finally they constrain the loop factors (c_gamma, c_g) = (1.18 +/- 0.12, 0.88 +/- 0.11) under the assumption of Standard Model tree-level couplings, the total Higgs width, and the invisible branching ratio, for which they quote 68% and 95% CL upper limits of 0.04 and 0.13 under Standard Model visible couplings.","tokens_in":13989,"tokens_out":4301,"duration_ms":73251,"significance":"The paper is a timely and careful update of a well-established program, and its two headline results are of real value to the community. The (M, epsilon) parameterization of Eq. (1) is a genuinely useful device: it converts the diffuse question of whether the Higgs resembles the Standard Model boson into two falsifiable numbers, the linearity of the mass dependence and the location of the scaling scale relative to v = 246 GeV, and the paper reports both with explicit chi-squared prices for the Standard Model point. The sign determination, if it holds under its stated assumptions, is a qualitative step beyond coupling-strength measurements, since it constrains the interference structure of the photon-photon loop and excludes the anti-dilaton branch c = -a. The analysis is also commendably candid about its inputs: Section 2 explicitly reports where the reconstructed likelihoods depart from the official combinations, and Section 3 explicitly states the assumption under which the sign discrimination is obtained. The results are derived from public data and are in principle reproducible, which is a strength worth naming.","major_comments":[{"comment":"The abstract states without qualification that 'Our global analysis of its couplings to fermions and massive bosons determines that they have the same relative sign as in the Standard Model', and the Conclusions repeat the claim in the same unconditional form. Section 3 and the right panel of Fig. 4 show that this is not what the analysis establishes: when c_gamma and c_g are marginalized, the paper itself reports that 'the symmetry between the solutions with c > 0 and < 0 is restored'. The sign result is therefore conditional on the assumption that no non-Standard-Model particles contribute to the H gamma gamma and H g g amplitudes, and that condition should appear in the abstract and in the Conclusions, not only in the body. As written, a reader of the abstract alone would take away a stronger result than the paper delivers.","section":"Abstract, sentence s2, and Section 6","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sign discrimination is a Delta chi-squared of about 9 obtained from a likelihood that the authors reconstruct themselves, and Section 2 documents that the reconstruction is not exact for the channel that drives it: the combined CMS photon-photon value is quoted as mu = 0.72 +0.24/-0.26 against the official 0.78 +0.28/-0.26. I do not think this discrepancy alone threatens the sign result, since the c < 0 branch predicts a photon-photon rate substantially above the Standard Model and both experiments sit at or below it, but the paper should demonstrate that rather than leave it to the reader. A short quantitative check would settle the matter: repeat the (a, c) fit with the official CMS photon-photon combination, and separately with the photon-photon and gluon-fusion theory uncertainties inflated to the 10% level, and report the resulting Delta chi-squared between the two sign branches. If the gap remains of order 9, the claim is robust and the check strengthens the paper; if it drops substantially, the reader deserves to know.","section":"Section 2 and Fig. 2, bottom right panel","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The fit treats the subchannels as statistically independent, and the paper does not discuss the correlations induced by common systematic uncertainties, in particular luminosity, jet energy scale, and the gluon-fusion theory uncertainty, which are shared across channels within an experiment and across experiments. Ignoring such correlations generally underestimates the uncertainties and can sharpen an apparent discrimination. Since the central claim rests on a likelihood difference, the paper should state explicitly whether any correlations were included, and if not, estimate their effect on the quoted errors in Eqs. (5) and (6) and on the sign gap. This is a methodological transparency issue rather than an error, but it bears directly on the headline number.","section":"Section 2, likelihood combination","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The (M, epsilon) fit is presented as a direct probe of the mass dependence, but the text does not state whether it assumes Standard Model loop content for c_gamma and c_g, as the (a, c) fit of Fig. 2 does. Given that the same data and the same sign ambiguity enter, and given that the right panel of Fig. 4 shows the loop assumptions matter, the paper should say explicitly what is held fixed in Fig. 6 and whether marginalizing over the loop factors changes the quoted ranges. The strongly asymmetric error on M, +20/-10 GeV, also warrants a sentence: it suggests a non-parabolic likelihood, and the reader should be told how the 68% range was extracted.","section":"Section 4, Eq. (7) and Fig. 6","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The text refers to 'the CERN Council in December 2013'; the paper is dated March 2013 and the reference is clearly to December 2012. The same sentence gives the ATLAS photon-photon result as 1.80+0.4-0.36, with inconsistent significant figures between the central value and its errors.","section":"Section 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Reference [18] gives 'arXiv:1303:1812' with a colon where a period belongs, and reference [16] contains an unrendered LaTeX command, 'JHEP bf 10 (2012) 062', which should read JHEP 10 (2012) 062.","section":"References","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption reads 'marginalizing over the invisible branching ration BRinv'; 'ration' should be 'ratio'.","section":"Fig. 10, right panel caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption names the experiments as 'ATLAS, CDF, D0 and CMS' while the body text refers throughout to the Tevatron combination. The caption should be self-contained and use one naming convention consistently.","section":"Fig. 1 caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The quoted errors on a and c should be accompanied by a statement of the convention used to extract them, for example the Delta chi-squared = 1 range after marginalizing over the other parameter, so that the numbers can be compared directly with other fits in the literature.","section":"Section 3, Eq. (5)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text explains the shading and contour convention once, in Section 3, but the figure captions do not repeat it. Each caption should state that lighter shading indicates more likely regions and that the dotted, dashed and solid contours correspond to 68, 95 and 99% CL, so that the figures stand on their own.","section":"Section 3, Figs. 2 and 3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The definition R_Vis = Gamma_Vis / Gamma_SM_Tot assumes that the visible width contains no non-Standard-Model contributions. The text should say so explicitly at the equation, since the subsequent limits on BR_inv are only as strong as that assumption, and the dot-dashed curve in Fig. 10, whose shallow minimum near BR_inv ~ 0.4 requires a > 1, deserves a sentence of interpretation.","section":"Section 5, Eq. (8)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The concluding remark that the measurements 'provide circumstantial support to supersymmetry as opposed to these minimal composite alternatives' is not derived from any fit presented in the paper. It is a reasonable opinion, but it should either be supported by a fit to a supersymmetric benchmark or be explicitly labelled as the authors' interpretation rather than a result of the analysis.","section":"Section 6","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is a solid, timely contribution and the analysis is standard and well documented. My main reservation is one of claim scope rather than correctness: the abstract and Conclusions state the sign determination unconditionally, while Section 3 and Fig. 4 show it depends on assuming no new physics in the photon-photon and gluon-gluon loops. The authors are aware of this and say so in the body, so the fix is editorial. The requested robustness check on the reconstructed CMS photon-photon likelihood is worth insisting on, since it is cheap and would put the headline number beyond dispute."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Quick read on Ellis–You 1303.3879. It is the Moriond-2013 update of their Higgs coupling fit, and the numbers people will quote are a = 1.03 ± 0.06, c = 0.84 ± 0.15, and a Δχ² ≈ 9 preference for c > 0 over the wrong-sign branch. The (M, ε) fit is the more interesting result to me: M = 244 +20/−10 GeV and ε = −0.022 +0.042/−0.021, i.e. the couplings are linear in mass to a couple of percent and normalized by v to about 10%. That is a clean, quotable statement, and the right panel of Fig. 6, where each species' coupling is predicted with that species dropped from the fit, is the figure I would actually show someone. The invisible-BR limit (4% at 68% CL under SM couplings) is a useful by-product. The (a, c) and (M, ε) machinery is theirs from ref. [17] and they say so up front, so the new content is the dataset and the resulting numbers, not a new method. That is fine for what this is. On the reader's worry about the CMS γγ reconstruction: I don't think it lands. Their 0.72 versus the official 0.78, with ±0.25 errors, is a quarter sigma in one channel. The c < 0 branch predicts a γγ rate roughly double the SM, so it is the qualitative absence of an enhancement in both ATLAS and CMS that kills it, not the central value. The genuine condition is the one the authors state in Section 3 and show in the right panel of Fig. 4: assume no new particles in the γγ and gg loops, and the sign is fixed; marginalize over cγ and, as they say themselves, the symmetry is restored. So the 'first time' claim is conditional, and the body is honest about it. My one real complaint is that the abstract states the sign result flat, with no mention of the loop assumption. That is the sentence a plenary speaker will read out. Minor: no likelihood code or digitized inputs, so reproducing the subchannel combinations means redoing them; and 'December 2013' for the CERN Council should be 2012. This is for anyone quoting post-Moriond Higgs coupling numbers, and the (M, ε) plot is the useful framing for 'is it really the Higgs.' Send it out; I would cite it.","headline":"A solid Moriond-2013 update of the Ellis–You Higgs coupling fit; the sign result is real but conditional on SM loop content, and the abstract hides that condition.","tokens_in":14717,"tokens_out":1490,"would_cite":true,"duration_ms":30036,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["14.80.Bn","12.60.Fr","13.85.Lg"],"model":"selfhosted/deepseek-v4.1-flash","headline":"A global fit to the first full round of LHC Higgs data finds that the new particle's couplings to fermions and to heavy bosons carry the same relative sign as in the Standard Model, and that their sizes track particle masses almost exactly.","keywords":["Higgs boson couplings","global fit","electroweak symmetry breaking","fermion sign","mass scaling","invisible Higgs width","composite Higgs","diphoton loop"],"falsifier":"An independent global fit using the official CMS and ATLAS diphoton likelihoods without subchannel reconstruction, or the same fit repeated with the later full 7+8 TeV diphoton datasets, would show whether the $c>0$ branch survives at $Delta chi^2 sim 9$ or collapses toward degeneracy.","tokens_in":14851,"feed_emoji":"⚛️","tokens_out":2678,"duration_ms":55155,"temperature":0.7,"pith_summary":"The paper asks a narrow but consequential question: the particle discovered at the LHC is a Higgs boson, but is it *the* Standard Model Higgs boson? To answer, the authors collect essentially all public Higgs measurements available in March 2013 — signal strengths in the $bbar b$, $tau^+tau^-$, $gammagamma$, $WW^*$ and $ZZ^*$ channels from ATLAS, CMS and the Tevatron — and fit them to a two-parameter model in which the couplings to massive bosons and to fermions are each rescaled by a single number, $a$ and $c$. The combined signal strength comes out at $mu = 1.02^{+0.11}_{-0.12}$, i.e. the overall rate is already within about 13% of the Standard Model at 68% confidence. The headline result is that the fit now prefers $c>0$ over $c<0$ by $Delta chi^2 sim 9$, fixing the relative sign of the fermion and boson couplings for the first time. A second, more physical parameterization asks whether the couplings really scale with mass as $m^{1+epsilon}$ and whether the scale of that scaling is the electroweak scale $v = 246$ GeV; the data say yes to both, to within a few percent and 10% respectively.","feed_headline":"Higgs couplings now match the Standard Model's sign","feed_subtitle":"A global fit to early 2013 LHC data finds fermion and boson couplings share the Standard Model's relative sign and scale with mass.","key_machinery":"The machinery is a global likelihood built channel by channel from published signal strengths, event counts, or 95% CL limits, then combined across experiments and subchannels — including separate 0-, 1- and 2-jet bins whose vector-boson-fusion content (2%, 12%, 81%) is what breaks degeneracies between production and decay couplings. That likelihood is then mapped onto two coupling parameterizations: the multiplicative $(a, c)$ rescaling of an effective Lagrangian in which $a$ and $c$ measure deviations from the Standard Model for bosons and fermions, and the mass-scaling $(M, epsilon)$ form that tests whether the couplings follow $m^{1+epsilon}$ and whether the scale is $v$. The 2-jet bins do specific work here: they exclude the band around $c sim 0$ where gluon-fusion production would fake the signal, which is what sharpens the sign discrimination. Named alternatives — the composite-Higgs MCHM4 line $a = c$, the MCHM5 curve, and the anti-dilaton line $c = -a$ — are drawn on the same plane so the reader can see which survive.","core_discovery":"Two claims carry the paper. First, the relative sign of the Higgs couplings to fermions and to massive vector bosons is now determined: the global minimum sits at $a = 1.03 pm 0.06$ and $c = 0.84 pm 0.15$, and the mirror branch with $c<0$ is disfavoured by $Delta chi^2 sim 9$. This is not a symmetry of the fit — it comes from interference between the top-quark and $W$-boson loops in $H to gammagamma$, so the newly reported CMS diphoton result, compatible with the Standard Model, is what tips the balance. Second, using the parameterization $lambda_f = sqrt{2}(m_f/M)^{1+epsilon}$ and $g_V = 2 m_V^{2(1+epsilon)}/M^{1+2epsilon}$, the authors find $M = 244^{+20}_{-10}$ GeV and $epsilon = -0.022^{+0.042}_{-0.021}$, with a likelihood penalty of only $Delta chi^2 = 0.12$ for the Standard Model point $(M, epsilon) = (246 text{ GeV}, 0)$. In other words the couplings are linear in mass to within a few percent and normalized by the conventional symmetry-breaking scale to within 10%. Subsidiary results: the loop-induced couplings come out at $c_gamma = 1.18 pm 0.12$ and $c_g = 0.88 pm 0.11$ when the tree-level couplings are held at their Standard Model values; the total Higgs width lies within 20% of the Standard Model at 68% confidence; and, assuming Standard Model couplings and no new loop contributions, the invisible branching ratio is bounded at 0.04 (68% CL) and 0.13 (95% CL).","pith_inferences":["The sign discrimination is a statement about loop interference, not about a direct measurement of a fermion coupling sign; it therefore tests the *relative* structure of the $t$-quark and $W$ contributions to $H to gammagamma$, and a future independent extraction of that loop (for example from off-shell or high-$p_T$ Higgs production) would be a genuinely separate check.","The $(M, epsilon)$ result suggests a natural next step: rather than fitting one global $(M, epsilon)$, fit separate exponents per fermion species, since the current data are already sensitive at the few-percent level and the $tau$ and $b$ channels will tighten fastest.","If the couplings continue to sit on the linear, $v$-normalized line as errors shrink, the parameter space of composite and dilaton-like alternatives shrinks toward the decoupling limit, which inverts the usual logic: the Higgs mass and the coupling line would together be circumstantial evidence for a weakly coupled sector rather than a strongly coupled one.","The invisible-width bound is derived under the assumption of no new loop contributions; relaxing that assumption, as the authors do in one variant, shifts the preferred region, so a future direct search for invisible Higgs decays would be the cleaner discriminant."],"forward_implications":["The anti-dilaton scenario $c = -a$, and any model in which fermion and boson couplings enter with opposite relative sign, is excluded at roughly the $3sigma$ level by the current combination.","The absence of a significant diphoton anomaly means there is no longer empirical pressure for extra charged particles in the $H gammagamma$ loop; any new physics there must be arranged to leave $c_g$ near unity, since $c_g = 0.88 pm 0.11$.","Minimal composite-Higgs models remain viable only near the Standard Model limit, so the data already constrain the compositeness scale to be well above the electroweak scale in those constructions.","Because the total width is pinned to within 20% of the Standard Model value, the room left for invisible or non-Standard Higgs decays is small: a few percent at 68% confidence under Standard Model assumptions."],"supporting_citations":[{"why":"Moriond 2013 electroweak session, the primary source of the ATLAS and CMS Higgs signal strengths the fit is built on.","marker":"[13]"},{"why":"Moriond 2013 QCD session, source of the updated CMS $H to gammagamma$ result that drives the fermion-sign preference.","marker":"[15]"},{"why":"Tevatron combined $H to bbar b, tau^+tau^-, WW^*, gammagamma$ best-fit signal strengths with 1-sigma errors, used as direct fit inputs.","marker":"[23]"},{"why":"CMS Moriond talk supplying the $tau^+tau^-$, $ZZ^*$ and diphoton subchannel central values and production-mode fractions.","marker":"[7]"},{"why":"Effective nonlinear Lagrangian for the symmetry-breaking sector, and the MCHM4/MCHM5 composite-Higgs relations between $a$ and $c$ that the fit confronts.","marker":"[33]"},{"why":"The authors' previous global analysis, which established the $(a,c)$ and $(M,epsilon)$ parameterizations being updated here.","marker":"[17]"},{"why":"ATLAS 0-, 1- and 2-jet $WW^*$ inputs, including the vector-boson-fusion fractions that make the 2-jet bins discriminating.","marker":"[27]"},{"why":"LHC Higgs Cross Section Working Group conventions for the $kappa_V$, $kappa_f$ rescaling framework the paper adopts and compares against.","marker":"[19]"}],"fun_headline_variants":["Higgs couplings share the Standard Model’s relative sign","Couplings scale linearly with mass to within a few percent","Early LHC data favor the Standard Model sign for Higgs couplings","Higgs width lies within 20% of the Standard Model at 68% CL","Invisible Higgs branching is bounded at 0.04 under Standard Model assumptions"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The sign preference rests on the authors' own reconstruction of the CMS diphoton likelihood, which they report as $mu = 0.72^{+0.24}_{-0.26}$ against the official $0.78^{+0.28}_{-0.26}$ and which they themselves say largely drives the result, so a modest change in that one channel's treatment could weaken the $Delta chi^2 sim 9$ separation.","fun_headline_variants_meta":{"raw":{"variants":["Higgs couplings share the Standard Model’s relative sign","Couplings scale linearly with mass to within a few percent","Early LHC data favor the Standard Model sign for Higgs couplings","Higgs width lies within 20% of the Standard Model at 68% CL","Invisible Higgs branching is bounded at 0.04 under Standard Model assumptions"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2468,"prompt_tokens":1073,"completion_tokens":1395,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":1316}},"tokens_in":1073,"tokens_out":1395,"duration_ms":23442,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":1316,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T07:53:43.159021+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"An independent global fit using the official CMS and ATLAS diphoton likelihoods without subchannel reconstruction, or the same fit repeated with the later full 7+8 TeV diphoton datasets, would show whether the $c>0$ branch survives at $Delta chi^2 sim 9$ or collapses toward degeneracy.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Moriond 2013 electroweak session, the primary source of the ATLAS and CMS Higgs signal strengths the fit is built on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Moriond 2013 QCD session, source of the updated CMS $H to gammagamma$ result that drives the fermion-sign preference."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Tevatron combined $H to bbar b, tau^+tau^-, WW^*, gammagamma$ best-fit signal strengths with 1-sigma errors, used as direct fit inputs."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"CMS Moriond talk supplying the $tau^+tau^-$, $ZZ^*$ and diphoton subchannel central values and production-mode fractions."},{"cited_title":"Strong Double Higgs Production at the LHC","cited_arxiv_id":"1002.1011","evidence_quote":"Effective nonlinear Lagrangian for the symmetry-breaking sector, and the MCHM4/MCHM5 composite-Higgs relations between $a$ and $c$ that the fit confronts."},{"cited_title":"Global Analysis of the Higgs Candidate with Mass ~ 125 GeV","cited_arxiv_id":"1207.1693","evidence_quote":"The authors' previous global analysis, which established the $(a,c)$ and $(M,epsilon)$ parameterizations being updated here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"ATLAS 0-, 1- and 2-jet $WW^*$ inputs, including the vector-boson-fusion fractions that make the 2-jet bins discriminating."},{"cited_title":"LHC HXSWG interim recommendations to explore the coupling structure of a Higgs-like particle","cited_arxiv_id":"1209.0040","evidence_quote":"LHC Higgs Cross Section Working Group conventions for the $kappa_V$, $kappa_f$ rescaling framework the paper adopts and compares against."}],"review_version":1}