Flavor-Violating Higgs and Top Decays in the Type 1B Flavorful Two-Higgs-Doublet Model with a Twist

Hao-Ran Ma; Jin-Lei Yang; Tai-Fu Feng; Ti-Bin Hou; Yu-Ju Peng; Zheng Li

REVIEW

Flavor-Violating Higgs and Top Decays in the Type 1B Flavorful Two-Higgs-Doublet Model with a Twist

Ti-Bin Hou, Zheng Li, Yu-Ju Peng, Jin-Lei Yang, Hao-Ran Ma, Tai-Fu Feng

High Energy Physics - Phenomenology · Submitted 1 Sep 2026 · Published version

Pith's verdict: Competent scan of an existing model, but the central numerical predictions are conditional on a rotation ansatz the authors themselves call an additional simplifying assumption, and they don't test alternatives.

T0 · AI review

How this review was made: An AI referee (deepseek/deepseek-v4-flash) read the paper on September 9, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 3 major objections · 8 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2609.00692 (version 3)

Pith IDNVRPG664, Pith's permanent record of this paper and its review

CitationHao-Ran Ma, Jin-Lei Yang, Tai-Fu Feng, Ti-Bin Hou, Yu-Ju Peng, Zheng Li. "Flavor-Violating Higgs and Top Decays in the Type 1B Flavorful Two-Higgs-Doublet Model with a Twist." arXiv:2609.00692 (2026). https://pith.science/paper/NVRPG664

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

You asked about 2609.00692. Here's my take.

What this paper does: it takes the Type 1B F2HDM from Altmannshofer–Maddock (2018), runs a 542-point constrained scan with updated Higgs, flavor, and EW precision constraints, and computes one-loop vertex corrections for h→sb, h→db, h→uc, t→ch, and t→uh. That set of one-loop numbers, the surviving-sample statistics, and the rare inversion points (0.55% where tree-loop interference flips the h→sb vs h→db hierarchy) are new. The paper is transparent about its cut-flow, conventions, and code stack. The internal consistency is fine.

Now the soft spots, in proportion. The biggest is the rotation ansatz of Eq. (17): V_Lu≈I, V_Ru≈I, V_Ld≈V_CKM, V_Rd≈I. The paper concedes in Sec. IIB that this is 'an additional simplifying assumption' not fixed by the texture. Under this assumption, the dominant entries in M'_d become −V*_ts m_b and −V*_td m_b, which gives the narrow CKM-controlled bands that the paper presents as stable predictions for h→sb and h→db. If the CKM were moved to the up sector or if V_Rd had small off-diagonals, the hierarchy would shift. The authors perform no sensitivity study. The claimed numerical hierarchy is therefore a conditional result, not a robust discovery.

Second, the 'one-loop-updated' quark rates omit counterterms and wave-function renormalization—the paper only includes one-loop vertex triangles. The authors are open about this, but the numerical quantiles (especially the three inversion points) should be read as illustrative. Third, the constraint set uses 2007 UTfit intervals and a relative Δχ² cut (Δχ²<1) on HiggsSignals, which is not a rigorous global fit. The surviving sample is useful but not definitive.

On the plus side, the qualitative pattern (down-sector CKM control, lepton and top alignment control, first-second generation suppression) is robust, and the paper states its caveats honestly. The h→μτ rate reaching 2.7e-4 is within HL-LHC reach, and the strong B→X_sγ correlation in the down-sector rates gives an external tie.

Who this is for: theorists working on flavorful 2HDMs or flavor-violating Higgs phenomenology. It's a solid but incremental update. A reader who needs the most up-to-date numbers for this specific model will find it useful.

Recommendation: deserves peer review. The referees should ask for a sensitivity study on the rotation ansatz and a note on the incompleteness of the loop calculation. With those additions, the paper would be publishable.

Referee report

Significance

The qualitative hierarchy pattern, if robust, gives concrete targets for flavor-violation searches in this model class: h→μτ as the experimentally cleanest channel, h→sb as the largest quark-flavor-violating Higgs mode (with the paper honestly noting its reconstruction difficulties), and t→ch as the leading top-flavor probe. Strengths of the manuscript include a reproducible machine-checked pipeline (SARAH/SPheno/HiggsSignals/FeynArts-FormCalc-LoopTools), an explicit cut-flow (Fig. 3), honest disclosure that the rotation ansatz of Eq. (17) is not fixed by the texture, explicit quantiles and hierarchy fractions, and a clear separation of tree-level and loop-updated claims. The central quantitative results are, however, conditional on two assumptions that the paper identifies but does not stress-test: the down-sector CKM rotation ansatz of Eq. (17), which fixes the leading entries of Eq. (15), and the triangle-only one-loop scheme of Eq. (25), which produces order-unity or larger relative corrections.

Major objections (3)

[Sec. IIB, Eq. (17); Sec. IVE, Table III] The narrow, CKM-controlled tree-level bands for BR(h→sb) and BR(h→db) are a central quantitative claim, but they follow directly from the rotation ansatz V_Lu≈I, V_Ru≈I, V_Ld≈V_CKM, V_Rd≈I of Eq. (17), which the authors concede in Sec. IIB is not fixed by the texture, with V_Rd≈I described as an additional simplifying assumption. Under this ansatz the leading entries of M'_d in Eq. (15) are the fixed terms −V*_ts m_b and −V*_td m_b; any alternative rotation consistent with the textures (for example, a non-negligible V_Rd with texture-level O(1) off-diagonal entries, or CKM mixing generated in the up sector) would redistribute flavor violation through Eqs. (11)-(13) and change both the magnitudes and the ordering of the leading channels. No sensitivity study of this choice is provided, so the claimed stability of the down-sector predictions is not established beyond the specific ansatz. I request a companion scan with at least two alternative rotation conventions and an explicit statement of which hierarchy claims survive.
[Sec. IIIC-IVD, Eq. (25); Table II; Sec. IVF] The one-loop updates include only the triangle-vertex diagrams of Figs. 1-2, with no external-leg self-energies, Yukawa counterterms, or scalar wave-function renormalization, and no renormalization-scheme discussion. Table II shows that the resulting relative corrections are order unity in the headline down-sector channels (median δ1L = −0.444 for h→sb and −0.422 for h→db) and enormous elsewhere (h→uc median 4.09×10^3 and maximum 4.25×10^10; t→ch maximum 3.06×10^4). In these regimes the loop contribution is not a small perturbation of the tree amplitude, and the omitted counterterm and self-energy pieces, which are generically of the same order as a complete one-loop vertex, cannot be assumed negligible. This bears directly on the hierarchy-inversion claim of Sec. IVF, since the three inversion points require a near-total cancellation (K_sb ≈ 7×10^-3), and both the location and the existence of such cancellations could shift under a complete renormalized computation. I request either the full one-loop renormalized calculation for the quoted channels, or a reframing of the quantitative claims to tree level with the vertex triangles presented as an indicative estimate accompanied by an estimate of the omitted terms.
[Sec. IVC, IVE, and IVH; Fig. 9] The interpretation of the down-sector rates as texture-level predictions is complicated by the selection itself. Figure 9 shows ρ_S(BR(B→X_sγ), log10 BR(h→sb)) = −0.909 and ρ_S(BR(B→X_sγ), log10 BR(h→db)) = −0.819, making BR(B→X_sγ) the dominant monotonic driver of both headline channels in the surviving sample. Since the B→X_sγ selection of Eq. (39) removes the high-BR(B→X_sγ) points, and the anti-correlation means these are precisely the low-BR(h→sb) and low-BR(h→db) points, the surviving distribution is compressed against the selection boundary; the reported medians and the 'narrow band' of Table III are therefore a mixture of the CKM-texture locking emphasized in Sec. IVE and constraint-edge selection. The paper should separate these two effects, for example by reporting the pull of the final sample with respect to the B→X_sγ interval or by recomputing the bands under a tighter and a looser B→X_sγ window, so that the prospective statements in Sec. IVH are not over-read as texture-determined predictions.

Minor objections (8)

[Sec. IVB, Fig. 4] Medians are quoted for six of the eight channels but not for h→eτ and h→eμ; please add them so that the abstract's relation BR(h→eτ)∼BR(h→eμ) can be checked against reported numbers.
[Eq. (44)] The typeset definition of O^f,ij_eff appears corrupted ('/radicaltp/radicalvertex/radicalvertex√'); it should read sqrt((|O^f_ij|^2+|O^f_ji|^2)/2).
[References] Several entries have incomplete bibliographic data, including missing years or journal numbers in Refs. [11], [12], [22], [25], [27]-[32], and [35]; please standardize all entries.
[Sec. IIB] The statement that the O-coefficients 'exactly reconstruct the physical fermion masses' is not accompanied by the reconstruction algorithm; please describe the fitting or iteration procedure, or point to the public code, so the texture scan is reproducible.
[Sec. IIID, Eq. (36)] Defining the Higgs-signal-strength cut as Δχ²≤1 with χ²_HS,min taken from the generated sample makes the selection self-referential; please state the HiggsSignals data set and version and the reference point so that the threshold 150.872 is reproducible, or use a fixed SM reference.
[Sec. IIIE and Fig. 3] The label 'HS/HB' in Fig. 3 is not defined in the text, and HiggsBounds is never mentioned; please clarify whether a HiggsBounds selection is part of the cut-flow and, if so, where it enters.
[Table V] For h→sb, h→db, and h→uc the 'Source' column lists Refs. [26,29], which are theory papers; please state explicitly that these are indirect, analysis-dependent constraints rather than direct experimental limits.
[Sec. IIIB] Please specify the numerical values of the total widths used for the branching-ratio normalization in Eqs. (29) and (31); the current wording ('the widths used in the corresponding numerical normalization') is vague.

How Pith checks papers → · Full audit of this paper → · Review history (2 revisions) →

The paper

The paper claims

The Type 1B two-Higgs-doublet model predicts that h→sb, t→ch, and h→μτ are the leading flavor-violating decay modes, with h→sb far above h→db and h→μτ far above the electron channels.

The reading

This paper argues that the Type 1B flavorful two-Higgs-doublet model — a two-doublet flavor structure in which one doublet gives mass mostly to the first two generations and the other to the third — predicts a strict, correlated pattern of rare Higgs and top decays. After applying Higgs signal-strength, electroweak-precision, meson-mixing, $B\to X_s\gamma$, and charged-lepton-flavor constraints to 50,000 parameter points, the surviving 542-point sample shows $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$, $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$, and $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau),\,\mathrm{BR}(h\to e\mu)$, with median one-loop-updated rates of $8.43\times10^{-4}$, $2.59\times10^{-5}$, $3.93\times10^{-8}$, and $5.41\times10^{-6}$ in the leading channels. The paper's mechanism is that the down-sector Higgs rates are locked to CKM factors and therefore narrow, while the top and muon-tau channels are controlled by alignment and texture entries. If these predictions hold, they identify the specific flavor-violating channels colliders should target, with $h\to\mu\tau$ the experimentally cleanest and $h\to sb$ the largest but hardest to reconstruct.

What carries the argument

The load-bearing machinery is the CKM-down mass reconstruction of the Type 1B Yukawa texture. The two doublets are assigned so that the first two generations couple dominantly to $\Phi_1$ and the third generation to $\Phi_2$, and the numerical rotation ansatz of Eq. (17) places all CKM mixing in the left-handed down-quark sector. In the down-sector mass matrix (15), the entries that drive $h\to sb$ and $h\to db$ are the fixed CKM-weighted terms $-V^*_{ts}m_b$ and $-V^*_{td}m_b$; scanned corrections are suppressed by $m_s/m_b$, which locks the tree-level rates into narrow bands and produces the down-sector hierarchy. The one-loop vertex corrections enter as tree-plus-triangle amplitudes, and the interference term between the two is the mechanism behind the rare $\mathrm{BR}_{1L}(h\to db)>\mathrm{BR}_{1L}(h\to sb)$ inversion points.

Extended reading notes

Core claim

The central claim is that the Type 1B F2HDM predicts a stable flavor-violating hierarchy: $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$, $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$, and $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau)\sim\mathrm{BR}(h\to e\mu)$ across the constrained parameter sample. In the 542 surviving points the median one-loop branching ratios are $8.43\times10^{-4}$ for $h\to sb$, $2.59\times10^{-5}$ for $h\to db$, $3.93\times10^{-8}$ for $t\to ch$, and $5.41\times10^{-6}$ for $h\to\mu\tau$, with maxima $1.91\times10^{-3}$, $7.20\times10^{-5}$, $2.32\times10^{-6}$, and $2.70\times10^{-4}$. The near-constancy of the down-sector rates follows from the mass-matrix entries $-V^*_{ts}m_b$ and $-V^*_{td}m_b$, which dominate the scanned texture corrections, whereas $t\to ch$ and $h\to\mu\tau$ spread over orders of magnitude with alignment and texture sensitivity. One-loop vertex corrections preserve the hierarchy in $539$ of $542$ points; in the remaining three, destructive tree–loop interference suppresses $h\to sb$ strongly enough that the one-loop $h\to db$ rate exceeds it. The paper concludes that $h\to sb$, $t\to ch$, $h\to db$, and $h\to\mu\tau$ are the channels a future experiment should probe in this model class.

Load-bearing premise

The load-bearing premise is the numerical rotation choice of Eq. (17) — all CKM mixing in the left-handed down-quark sector and no right-handed mixing — which the paper itself calls an additional simplifying assumption; the Yukawa texture does not fix this rotation, so a different valid choice would redistribute the flavor violation and change which decay dominates.

Editorial extensions

If this is right

A collider program for this model class should prioritize $h\to sb$, $h\to\mu\tau$, and $t\to ch$: the first is numerically the largest quark-flavor-violating mode, the second is the cleanest lepton mode, and the third is the leading top-flavor probe.
The correlated pair $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$ and $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$ gives a pattern of ratios that can distinguish this two-doublet flavor structure from structures with differently placed CKM mixing.
Because $\mathrm{BR}(h\to sb)$ and $\mathrm{BR}(h\to db)$ are tightly anticorrelated with $\mathrm{BR}(B\to X_s\gamma)$, better measurements of the radiative B decay directly sharpen the model's Higgs-rate predictions.
One-loop corrections preserve the overall hierarchy in $539$ of $542$ surviving points; the rare inverted points are a genuine destructive tree–loop interference effect rather than evidence against the model's ordering.
The surviving sample keeps $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau),\mathrm{BR}(h\to e\mu)$, so observing the electron channels at rates comparable to $h\to\mu\tau$ would already point away from this realization.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Inference: The rotation ansatz of Eq. (17) is the natural first thing to vary; the paper concedes the texture does not fix the rotation, so moving the CKM matrix to the up sector could redistribute the same flavor violation and change which channel is largest.
Inference: The strong anticorrelation of the surviving down-sector rates with $\mathrm{BR}(B\to X_s\gamma)$ suggests that a single combined fit of radiative B decay, meson mixing, and Higgs FCNC rates could identify which scalar couplings the $B\to X_s\gamma$ bound is actually constraining.
Inference: Because the paper stops at rate-level predictions, a dedicated strange-tagging search for $h\to sb$ at the $10^{-3}$ level is the concrete detector-level test that the results motivate but do not carry out.

Original abstract

We study flavor-violating decays of the SM-like Higgs boson and rare top decays in the Type 1B realization of the flavorful two-Higgs-doublet model with a twist, focusing on $h\to sb$, $h\to db$, $h\to uc$, $h\to\mu\tau$, $h\to e\tau$, $h\to e\mu$, $t\to ch$, and $t\to uh$. We present the relevant amplitude and branching-ratio formulae, include one-loop vertex corrections, and impose constraints from Higgs signal-strength measurements, electroweak precision tests, meson mixing, $B\to X_s\gamma$, direct HFV searches and CLFV constraints, including $\mu-e$ conversion. The surviving parameter space favors $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$, $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$, and $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau)\sim\mathrm{BR}(h\to e\mu)$ in the lepton sector. One-loop effects usually preserve these hierarchies, but can modify individual channels through different mechanisms: sizable destructive tree--loop interference can reduce $h\to sb$ enough to yield rare $\mathrm{BR}_{\mathrm{1L}}(h\to db)>\mathrm{BR}_{\mathrm{1L}}(h\to sb)$ points, whereas the large relative corrections found in rare top decays mainly reflect strong SM-like alignment suppressing the tree-level amplitudes.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-09-23: 1 request, 4,302 input tokens.

What the abstract itself states, as probabilities: claims first 0.02 · states limitation 0.55 · human participants 0.01 · controlled comparison 0.07 · randomized assignment 0.01 · states quantitative result 0.18 · compares with prior results 0.30 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v2. Results and the main claim are underlined; show every role.

We study flavor-violating decays of the SM-like Higgs boson and rare top decays in the Type 1B realization of the flavorful two-Higgs-doublet model with a twist, focusing on $h\to sb$, $h\to db$, $h\to uc$, $h\to\mu\tau$, $h\to e\tau$, $h\to e\mu$, $t\to ch$, and $t\to uh$. We present the relevant amplitude and branching-ratio formulae, include one-loop vertex corrections, and impose constraints from Higgs signal-strength measurements, electroweak precision tests, meson mixing, $B\to X_s\gamma$, direct HFV searches and CLFV constraints, including $\mu-e$ conversion. The surviving parameter space favors $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$, $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$, and $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau)\sim\mathrm{BR}(h\to e\mu)$ in the lepton sector. One-loop effects usually preserve these hierarchies, but can modify individual channels through different mechanisms: sizable destructive tree--loop interference can reduce $h\to sb$ enough to yield rare $\mathrm{BR}_{\mathrm{1L}}(h\to db)>\mathrm{BR}_{\mathrm{1L}}(h\to sb)$ points, whereas the large relative corrections found in rare top decays mainly reflect strong SM-like alignment suppressing the tree-level amplitudes.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	We study flavor-violating decays of the SM-like Higgs boson and rare top decays in the Type 1B realization of the flavorful two-Higgs-doublet model with a twist, focusing on $h\to sb$, $h\to db$, $h\to uc$, $h\to\mu\tau$, $h\to e\tau$, $h\to e\mu$, $t\to ch$, and $t\to uh$.	contribution 1.00	0.89	0.05	asserted 0.77
s1	We present the relevant amplitude and branching-ratio formulae, include one-loop vertex corrections, and impose constraints from Higgs signal-strength measurements, electroweak precision tests, meson mixing, $B\to X_s\gamma$, direct HFV searches and CLFV constraints, including $\mu-e$ conversion.	method 0.86	0.93	0.03	asserted 0.97
s2	The surviving parameter space favors $\mathrm{BR}(h\to sb)\gg\mathrm{BR}(h\to db)$, $\mathrm{BR}(t\to ch)\gg\mathrm{BR}(t\to uh)$, and $\mathrm{BR}(h\to\mu\tau)\gg\mathrm{BR}(h\to e\tau)\sim\mathrm{BR}(h\to e\mu)$ in the lepton sector.	result 1.00	0.90	0.72	asserted 0.97
s3	One-loop effects usually preserve these hierarchies, but can modify individual channels through different mechanisms: sizable destructive tree--loop interference can reduce $h\to sb$ enough to yield rare $\mathrm{BR}_{\mathrm{1L}}(h\to db)>\mathrm{BR}_{\mathrm{1L}}(h\to sb)$ points, whereas the large relative corrections found in rare top decays mainly reflect strong SM-like alignment suppressing the tree-level amplitudes.	interpretation 0.53	0.86	0.15	asserted 0.54

What kind of paper the classifier reports this is: theoretical result 0.98 · evidence from simulation or computation 0.73.

What was examined, and what was not

sources read

arXiv listing abstract v2

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures

**Figure 1.** FIG. 1: One-loop triangle categories for the Higgs FCNC channels. The notation is

**Figure 2.** FIG. 2: One-loop triangle categories for the top FCNC channels

**Figure 3.** FIG. 3: Cut-flow for the Type 1B F2HDM numerical scan after the Higgs, flavor, and electroweak

**Figure 4.** FIG. 4: Overview of the flavor-violating rates for the eight channels, shown as

**Figure 5.** FIG. 5: Hierarchy-ratio distributions in the Type 1B F2HDM sample. The dashed black vertical line marks

**Figure 6.** FIG. 6: Fractions of points satisfying the down-sector, lepton-sector, and top-sector hierarchy conditions.

**Figure 7.** FIG. 7: Spearman-rank driver analysis for the main flavor-violating rates.

**Figure 8.** FIG. 8: Dependence of the main flavor-violating rates on

**Figure 10.** illustrates this denominator effect in more detail. In panel (a), the largest values of δ1L appear close to the alignment region. Panel (b) shows that the same region contains very small tree-level t → ch branching ratios. The corresponding one-loop-updated branching ratio is shown in

**Figure 11.** FIG. 11: Spearman correlations for representative tree-level Higgs and top FCNC branching ratios in the

**Figure 13.** FIG. 13: Interference-cosine projection of the rare hierarchy inversion. Panel (a) shows the full

**Figure 14.** FIG. 14: Comparison of the tree-level and SPheno one-loop branching ratios for the lepton-sector HFV

**Figure 15.** FIG. 15: Representative contributions to

Figures reproduced from arXiv:2609.00692 by the authors.

References

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.