{"id":"9f49fd6a-a3b3-4ef2-85a4-5d082cd5f3d1","arxiv_id":"2204.05296","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Non-degenerate top squark soft masses, and a Z-prime made leptophobic by mixing with vector-like leptons, are the two scenarios that can accommodate the CDF II W mass without conflicting with LHC searches.","lead":"Three electroweak global fits and four model classes are used to test whether new physics can produce the W mass reported by CDF II, about 76 MeV above the Standard Model prediction. The paper finds that only tuned or loop-suppressed scenarios survive LHC searches, with non-degenerate top squark soft masses giving the most viable fit.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The printed non-degenerate stop loop functions in Eq. (44) diverge as rr→1, contradicting the paper's own Fig. 7 caption reference values and the degenerate limit Eq. (42); the Fig. 8 allowed region built on them is not established as printed.","rationale":"The reader correctly identifies that the stop result is the paper's strongest specific claim and that it rests on unverified approximations. However, the reader's weakest_assumption focuses on the vacuum stability bounds Eq. (45)-(46), which the authors themselves describe as 'empirical and approximated' and which are therefore an acknowledged limitation rather than a new finding. The more concrete and load-bearing issue is the loop function in Eq. (44): as printed, it diverges at rr=1, contradicting both the degenerate limit Eq. (42) and the paper's own Fig. 7 caption reference values. This is a checkable internal inconsistency in the formula that generates the paper's headline stop result. It does not change the overall verdict category (CONDITIONAL remains appropriate for an exploratory 'Speculations' paper with several independent results), but it adds a specific condition: the Eq. (44) loop functions must be verified before the Fig. 8 allowed region is treated as established. The reader's other observations (the 6 MeV theory error is a hand-chosen input, no code shipped, the Fig. 8 caption mislabels the plotted scenario as 'degenerate' when it is non-degenerate) are correct but secondary. The paper's EW fit machinery is standard and cross-checked against Ref. [18], and the negative result for the left-handed SU(2)xSU(2) model is reported honestly; those parts of the paper hold up. The concern is localized to the non-degenerate stop section, which is the paper's most quantitative BSM claim.","tokens_in":23225,"tokens_out":15645,"duration_ms":391223,"concrete_test":"Take the analytic rr→1 limit of each term in Eq. (44) and compare to Eq. (42) and the Fig. 7 caption reference values. If the limit diverges (as the printed formula does), the formula is wrong and Fig. 7/Fig. 8 must be recomputed; if it matches the stated finite values, the concern is void.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's strongest result is that non-degenerate stop soft masses (rr = m_U3^2/m_Q3^2 < 1) enhance c_T, c_B, c_WB enough to open an allowed region in Fig. 8. That result rests entirely on the loop functions in Eq. (44). As printed, those functions do not reduce to the degenerate limit Eq. (42) at rr=1; they diverge. Expanding at rr=1-ε: the c_T X_t^2 coefficient goes as -7/(4ε^3), the c_T X_t^4 coefficient as -9/ε^4, the c_W X_t^2 coefficient as -4/ε^4, the c_B X_t^2 coefficient as 1/ε^3, and the c_WB X_t^2 coefficient as -3/(4ε^3). Yet the paper's own Fig. 7 caption states finite rr=1 reference values (1/40, -1/16, 1/40, 1/40, 1/30), and Eq. (42) gives the degenerate limit. So either the printed Eq. (44) contains a typo in the denominators or the enhancement factors in Fig. 7 and the allowed region in Fig. 8 are wrong. This is not the same as the reader's concern about the vacuum stability bounds, which the authors explicitly acknowledge as 'empirical and approximated'; the issue here is the loop function itself, which is presented as an exact one-loop result and is not flagged as approximate. If the printed formula is correct, the enhancement claim is unsupported; if it is a typo, the quantitative Fig. 7 and Fig. 8 results need to be recomputed with the corrected function.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript performs three electroweak global fits in the SMEFT framework — S-T, S-T-δG_F, and an eight-parameter flavor-universal fit — to quantify the tension between the CDF II m_W measurement and the PDG world average, and then surveys three classes of BSM models (extended gauge groups with Z'/W', vector-like top partners, and top squarks) for compatibility with the fit. The headline novel claim is that non-degenerate stop soft masses (rr = m_U3²/m_Q3² < 1) enhance the one-loop Wilson coefficients c_T, c_B, and c_WB enough to open an allowed region in Fig. 8 that satisfies LHC stop mass bounds and approximate vacuum stability constraints.","tokens_in":23796,"tokens_out":22008,"duration_ms":506525,"significance":"The EW fit framework is standard, uses a stated operator basis (Table III) and stated observables (Eqs. 2 and 3), and is cross-checked against Ref. [18] for the S and T results, which is a genuine strength. The model survey is a useful map of which TeV-scale explanations survive current LHC constraints, and the paper makes falsifiable predictions for future colliders (ILC, CEPC, FCC-ee, muon colliders) and for LHC searches. The non-degenerate stop result, if correct, is a concrete and testable TeV-scale explanation of a positive T parameter. The paper is transparent about several approximations, including the 6 MeV effective theory error, the CDF-alone versus PDG comparison, and the approximate vacuum stability bounds.","major_comments":[{"comment":"The printed non-degenerate loop functions do not reduce to the degenerate limit Eq. (42) as rr→1; they diverge. For example, the c_T coefficient of X_t²/m_Q3² contains (-1-5rr-2rr²)/(8(1-rr)³) + 3rr²/(4(1-rr)⁴) log rr, whose numerator at rr=1 is -8 while the denominator vanishes as (1-rr)³, giving a leading -1/ε³ divergence (with rr = 1-ε); the log term contributes -3/(4ε³), so the sum diverges as -7/(4ε³). The same happens for the X_t⁴ term and for c_W and c_B. Yet the Fig. 7 caption quotes finite rr=1 reference values (1/40, -1/16, 1/40, 1/40, 1/30), and Eq. (42) gives the degenerate limit. Either the printed Eq. (44) contains a typo in the denominators or the enhancement factors in Fig. 7 and the allowed region in Fig. 8 are wrong. Because the central claim of §III C (abstract:s9) rests on these functions, the authors should verify the loop functions, confirm they reduce to Eq. (42) at rr=1, and recompute Fig. 7 and Fig. 8 if a correction is needed. As a secondary point, the rr=1 reference value quoted in the Fig. 7 caption for c_T(X_t²/m_Q3²) is -1/16, whereas Eq. (42) gives -1/8; the normalization should be reconciled.","section":"§III C, Eq. (44) and Fig. 7","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The boundary of the viable region in Fig. 8 is set by two approximate vacuum stability constraints that the authors describe as 'empirical and approximated' and note depend on 'many other parameters.' The authors acknowledge this and show both bounds, which is appropriate. What remains unresolved is whether a full tunneling calculation would leave any viable region at the large X_t, small rr corner where the enhancement operates; a numerical check, or at least a statement of how the viable region shrinks under the more conservative bound, would strengthen the claim that the non-degenerate stop explanation is 'compatible with current constraints.'","section":"§III C, Fig. 8 and Eqs. (45)-(46)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The caption says 'with degenerate stop soft masses,' but the benchmarks shown are non-degenerate (rr = 0.9, 0.8, 0.6); the caption should be corrected.","section":"Fig. 8 caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption describes the gray shaded region as 'tachyonic light stop mass,' while the text calls it the 'color-breaking parameter space'; these should be reconciled.","section":"Fig. 8 caption and §III C text","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The SM row lists three numbers (31, 62, 76) under two χ² columns; the column structure should be clarified so the reader can tell which number corresponds to which scenario.","section":"Table I","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The printed coefficient '12c2w' is ambiguous; the exact normalization of δm_W in terms of T̂, Ŝ, and δG_F should be stated explicitly.","section":"Eq. (4)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The expression 'mZ′ = mW′ = 12√g1²+g2² v_∆' should read (1/2)√(g1²+g2²) v_∆; the factor should be corrected.","section":"Eq. (10)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The 6 MeV effective theory error added in quadrature to the m_W experimental uncertainty is a modeling choice that affects the fit results; a brief sensitivity study, or a justification of the value beyond the citation to Ref. [2], would strengthen the robustness of the quoted shifts.","section":"§II","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The comparison uses the CDF II measurement alone versus the PDG world average rather than a proper combination; the authors acknowledge this, but a sentence on how a combination would shift the fit results would help the reader.","section":"§II","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The claim that vector-like lepton mixing can eliminate the LHC di-muon bounds on Z' is sketched via Eq. (32) but not demonstrated with a scan; a figure showing the viable {tanφ, θ_L, θ_R} region would strengthen the realization claim in the abstract.","section":"§III A, Eq. (32)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The statement 'we can safely discard this possibility' for degenerate stops is stronger than the EFT-validity caveat at m_t̃ ~ 200-400 GeV warrants; softening the wording would be more accurate.","section":"§III C","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Several typographical errors should be corrected: 'simutenously' (should be 'simultaneously'), 'Lagarange' (should be 'Lagrangian'), 'unexclueded' (should be 'unexcluded'), 'parathensis' (should be 'parentheses'), 'dominants' (should be 'dominates'), and 'specifies' (should likely be 'sfermions').","section":"Throughout","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision","confidential_remarks":"The EW fit section is sound and cross-checked, and the model survey is a useful contribution. The top squark section is the most novel but also the most fragile: the printed Eq. (44) diverges at rr=1 and is inconsistent with both Eq. (42) and the Fig. 7 caption, so the headline result of §III C cannot be assessed as printed. This should be resolved before publication; if it is a typo, the corrected Fig. 7 and Fig. 8 should be provided."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Worth your time for the fit tables, not for the stop section as printed. The durable content is Section II: the S-T, S-T-δG_F, and eight-parameter fits redone with the CDF II m_W in place of the PDG average, with the operator basis and observable list stated explicitly and the S, T numbers cross-checked against the Gfitter values. The qualitative conclusion — you need positive T and positive δG_F with S near zero to move m_W without disturbing the Z pole — is correct and cleanly derived from Eq. (4). Table I is a genuinely handy summary. I also give them credit for reporting the negative result: the left-handed SU(2)₁×SU(2)₂ case makes the fit worse, and they say so rather than burying it. The observation that rr = m_U3²/m_Q3² < 1 enhances the custodial-breaking coefficients is a real extension of the degenerate-stop analysis, and the physical picture — a light right-handed stop with large left-right mixing — is sensible and less LHC-constrained than the degenerate case. That much is right. The problem is Eq. (44). Expand the printed c_T, c_W, c_B, c_WB coefficients at rr = 1 − ε and they diverge: −9/ε⁴ in the c_T X_t⁴ term, −7/(4ε³) in the c_T X_t² term, and so on. Eq. (42) gives a finite degenerate limit, and the Fig. 7 caption quotes finite rr = 1 reference values (1/40, −1/16, 1/40, 1/40, 1/30). So Eq. (44) contradicts both. My guess is a transcription slip in the (1−rr) powers — the Huo and Drozd-Ellis-Quevillon-You expressions are finite — but as printed, Fig. 7 and the Fig. 8 allowed region cannot be reproduced from the formula, and the enhancement claim is unsupported until someone recomputes. That is the one thing a referee must force. Smaller items: the Fig. 8 caption says 'degenerate' where the text means non-degenerate; the 6 MeV theory error is hand-chosen and never varied, which sets the required new-physics size; and the Z′ escape route works by tuning the fermion couplings to zero, which is evasion rather than explanation. The vacuum-stability caveat they flag themselves, so I won't hold it against them. Send it out. The fit material deserves refereeing and the stop section needs the correction before anyone builds on it.","headline":"A useful, honest post-CDF-II survey whose electroweak fit tables will be cited, but the top-squark headline result rests on a loop function that does not survive a check at rr = 1.","tokens_in":24441,"tokens_out":2580,"would_cite":true,"duration_ms":53280,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["12.60.-i","14.70.Fm","14.80.Ly"],"model":"selfhosted/deepseek-v4.1-flash","headline":"The CDF II W boson mass sits about seven sigma above the Standard Model prediction, and this paper asks which kinds of new physics could produce that shift without disturbing the Z-pole observables that already agree with theory.","keywords":["W boson mass","electroweak precision observables","Standard Model effective field theory","top squark","Z prime boson","vector-like quarks","CDF II","custodial symmetry"],"falsifier":"A dedicated calculation of the tunneling rate to the color-breaking vacuum across the large-$X_t$, small-$r_r$ corner of the stop fits, or a direct LHC search for a light right-handed stop with large left-right mixing in that region, would settle the claim; a future lepton-collider $m_W$ measurement near 1 MeV that confirms the Standard Model value would remove the anomaly it explains.","tokens_in":24546,"feed_emoji":"⚛️","tokens_out":5638,"duration_ms":150195,"temperature":0.7,"pith_summary":"The CDF II experiment measured the W boson mass at $80433.5$ MeV, roughly seven sigma above what a global electroweak fit predicts in the Standard Model. This paper asks what kind of new physics could produce that shift without spoiling the many Z-pole observables that already match the Standard Model, and then checks the candidate models against collider data. The authors first redo the electroweak fit in three effective-field-theory frameworks of increasing generality, and show that the data pull the fit toward a positive $T$ parameter together with a positive shift in the Fermi constant, while leaving $S$ essentially alone. They then survey three model classes: an extra $W'/Z'$ triplet, vector-like top partners, and top squarks. The first two are squeezed by direct LHC searches, while a top squark with unequal left- and right-handed soft masses survives, so the W mass excess would point to a light right-handed stop with large left-right mixing at the TeV scale.","feed_headline":"A lopsided top squark could absorb the CDF W mass excess","feed_subtitle":"Unequal left- and right-handed stop soft masses lift the right operators enough to fit a seven-sigma shift without breaking LHC bounds.","key_machinery":"The load-bearing device is the three-parameter electroweak fit in $S$, $T$ and $T$, where $T$ denotes the shift in the inferred Fermi constant from a four-fermion operator affecting muon decay. Only two independent combinations, $T - T$ and $S$, enter the Z-pole observables, so the fit has a flat direction that raises $m_W$ at fixed Z physics. That direction is what every model in the paper is scored against. The second piece of machinery is the one-loop matching of the top squark sector onto the same operator basis: the coefficients $c_T$, $c_W$, $c_B$ and $c_{WB}$ are written as explicit functions of the soft masses, the left-right mixing $X_t$ and $r_r$, and the paper shows that $r_r < 1$ turns the $X_t^4$ and $X_t^2$ terms up by a factor of a few relative to the degenerate case.","core_discovery":"The paper's central claim is that a top squark whose left- and right-handed soft masses differ can account for the CDF II W mass while staying inside existing LHC and vacuum-stability bounds, whereas the degenerate-mass version cannot. The argument runs through the electroweak fit. With the PDG world average for $m_W$, the three-parameter fit gives $T = 0.081$ and $T = 0.254$ once the CDF II value is used instead, with $T$ quoted at one $\\sigma$ as $0.081$ versus $0.254$, and the Fermi-constant shift moving from $(2.1)$ to $(15.7)$ in units of $10^{-4}$, while $S$ stays at $0.037$. Because the Z-pole data depend only on the combinations $T - T$ and $S$, a shift along the direction where the $T$ parameter and the Fermi-constant shift are equal raises $m_W$ while leaving the Z pole untouched. Degenerate stop soft masses generate the required one-loop Wilson coefficients only for stops of 200 to 400 GeV, below the LHC reach. Taking the soft-mass ratio $r_r = m^2_{U3}/m^2_{Q3} < 1$ enhances $c_T$, $c_B$ and $c_{WB}$ by factors of a few, opening an allowed region at large left-right mixing $X_t/m_{Q3} > 1$ with the left-handed soft mass in the TeV range.","pith_inferences":["The flat direction the paper identifies is effectively a one-number target, which means the model survey is less a search over theories than a test of which theories can hit that number without leaving other traces; the leptophobic $Z'$ construction is doing more work than the abstract's phrasing suggests.","Because the tension shrinks as the fit gains parameters, a reader should treat the seven sigma as framework-dependent rather than as a fixed experimental number, and the paper's own caveat that a new world average is needed makes the size of the anomaly a moving target.","If the non-degenerate stop corner is real, the same large $A_t$ that fixes $m_W$ also controls stop decays and Higgs-to-stop-pair phase space, so a stop search and a Higgs invisible-width measurement would probe the same parameter.","The paper's emphasis that the lightest stop should be right-handed to evade LHC bounds suggests a concrete search strategy: soft displaced vertices from a compressed right-handed stop spectrum, which the authors note is not yet covered."],"forward_implications":["If the CDF II value is correct, any viable model must shift the $T$ parameter and the Fermi constant together, so it needs both a source of custodial-symmetry breaking and a four-fermion operator that touches muon decay.","A $Z'$ explanation survives the LHC dilepton limits only if the $Z'$ is made fermiophobic or leptophobic, which requires extra vector fermions mixing with the Standard Model ones and gives exotic-fermion searches a target.","Vector-like top partners supply only part of the needed shift, so they work as one ingredient of an explanation rather than a complete one.","The preferred stop region has large left-right mixing, which also lifts the Higgs mass toward 125 GeV and ties the W mass explanation to a Higgs-sector prediction.","A $WW$-threshold scan at a future lepton collider, measuring $m_W$ to about 1 MeV, would establish whether the anomaly is real and would put an upper bound on the scale of the new physics.","The eight-parameter flavor-universal fit lowers the minimum $\\chi^2$ from 28.3 to 17.1, so part of the nominal seven-sigma tension can be absorbed by operators unrelated to $m_W$."],"supporting_citations":[{"why":"Supplies the CDF II W mass measurement that the entire paper interprets.","marker":"[1]"},{"why":"Supplies the PDG world-average $m_W$, the input parameters $m_Z$, $G_F$ and the branching-ratio data used in the fits.","marker":"[2]"},{"why":"Supplies the LEP Z-pole dataset and correlations that constrain the fit alongside $m_W$.","marker":"[9]"},{"why":"Supplies the Standard Model prediction and the 6 MeV fit uncertainty treated as an effective theory error.","marker":"[18]"},{"why":"Defines the hatted oblique parameters $S$ and $T$ used throughout the electroweak fit.","marker":"[19]"},{"why":"Supplies the ATLAS vector-like quark mass exclusions of 1.31 TeV for the singlet and 1.37 TeV for the doublet top partner.","marker":"[35]"},{"why":"Supplies the CMS dilepton resonance limit that excludes the $Z'$ best-fit region below 5 TeV.","marker":"[31]"},{"why":"Supplies the complete one-loop sfermion matching used to write the stop Wilson coefficients for non-degenerate soft masses.","marker":"[40]"},{"why":"Supplies the first approximate color-breaking vacuum bound, Eq. (45), that carves the allowed stop region.","marker":"[42]"},{"why":"Supplies the second approximate vacuum-stability bound, Eq. (46), shown as the dotted orange line in the stop fits.","marker":"[43]"}],"fun_headline_variants":["Split stop soft masses can fit the CDF W mass excess","Unequal stop soft masses open an allowed CDF W mass fit","Non-degenerate stop masses survive LHC bounds in CDF W fit","Top squark soft-mass split eases the CDF W mass tension"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The edge of the allowed stop region is drawn with two approximate vacuum-stability conditions that the authors themselves describe as empirical and approximated, and the paper does not independently check that they hold in the large-mixing, small-soft-mass-ratio corner where the explanation lives.","fun_headline_variants_meta":{"raw":{"variants":["Split stop soft masses can fit the CDF W mass excess","Unequal stop soft masses open an allowed CDF W mass fit","Non-degenerate stop masses survive LHC bounds in CDF W fit","Top squark soft-mass split eases the CDF W mass tension"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1752,"prompt_tokens":1122,"completion_tokens":630,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":565}},"tokens_in":1122,"tokens_out":630,"duration_ms":11793,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":565,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-30T22:46:38.799917+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"A dedicated calculation of the tunneling rate to the color-breaking vacuum across the large-$X_t$, small-$r_r$ corner of the stop fits, or a direct LHC search for a light right-handed stop with large left-right mixing in that region, would settle the claim; a future lepton-collider $m_W$ measurement near 1 MeV that confirms the Standard Model value would remove the anomaly it explains.","supporting_citations":[{"cited_title":"High-precision measurement of the W boson mass with the CDF II detector,","cited_arxiv_id":null,"evidence_quote":"Supplies the CDF II W mass measurement that the entire paper interprets."},{"cited_title":"Review of Particle Physics,","cited_arxiv_id":null,"evidence_quote":"Supplies the PDG world-average $m_W$, the input parameters $m_Z$, $G_F$ and the branching-ratio data used in the fits."},{"cited_title":"Precision Electroweak Measurements on the Z Resonance","cited_arxiv_id":"hep-ex/0509008","evidence_quote":"Supplies the LEP Z-pole dataset and correlations that constrain the fit alongside $m_W$."},{"cited_title":"The global electroweak fit at NNLO and prospects for the LHC and ILC","cited_arxiv_id":"1407.3792","evidence_quote":"Supplies the Standard Model prediction and the 6 MeV fit uncertainty treated as an effective theory error."},{"cited_title":"Electroweak symmetry breaking after LEP1 and LEP2","cited_arxiv_id":"hep-ph/0405040","evidence_quote":"Defines the hatted oblique parameters $S$ and $T$ used throughout the electroweak fit."},{"cited_title":"Combination of the searches for pair-produced vector-like partners of the third-generation quarks at $\\sqrt{s} =$ 13 TeV with the ATLAS detector","cited_arxiv_id":"1808.02343","evidence_quote":"Supplies the ATLAS vector-like quark mass exclusions of 1.31 TeV for the singlet and 1.37 TeV for the doublet top partner."},{"cited_title":"Search for resonant and nonresonant new phenomena in high-mass dilepton final states at $\\sqrt{s} = $ 13 TeV","cited_arxiv_id":"2103.02708","evidence_quote":"Supplies the CMS dilepton resonance limit that excludes the $Z'$ best-fit region below 5 TeV."},{"cited_title":"Effective Field Theory of Integrating out Sfermions in the MSSM: Complete One-Loop Analysis","cited_arxiv_id":"1509.05942","evidence_quote":"Supplies the complete one-loop sfermion matching used to write the stop Wilson coefficients for non-degenerate soft masses."},{"cited_title":"Phase Transitions and Vacuum Tunneling Into Charge and Color Breaking Minima in the MSSM","cited_arxiv_id":"hep-ph/9602414","evidence_quote":"Supplies the first approximate color-breaking vacuum bound, Eq. (45), that carves the allowed stop region."},{"cited_title":"Vacuum Stability and the MSSM Higgs Mass","cited_arxiv_id":"1310.4174","evidence_quote":"Supplies the second approximate vacuum-stability bound, Eq. (46), shown as the dotted orange line in the stop fits."}],"review_version":1}