Search for Higgs boson decay to a charm quark-antiquark pair via $t\bar{tH}$ production

Sebastian Wuchterl (on behalf of the CMS Collaboration)

REVIEW 1 major objections 3 minor 17 references

Search for Higgs boson decay to a charm quark-antiquark pair via $t\bar{tH}$ production

T0 review · 1 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A search for Higgs boson decays to charm quarks in $t\bar{t}H$ production, combined with earlier $VH$ data, sets the most stringent limit on the charm Yukawa coupling: observed $|\kappa_c|<3.5$ at 95% confidence.

desk verdict Solid incremental CMS result giving the tightest kappa_c bound to date; the headline claim needs a sharper caveat about fixed vs profiled kappa_b. read the letter →

arxiv 2506.02163 v1 pith:TPR6GRNZ submitted 2025-06-02 hep-ex

Sebastian Wuchterl (on behalf of the CMS Collaboration) This is my paper

classification hep-ex

keywords HiggsbosoncharmYukawacouplingHtoccttproductionCMSkappa_cjettaggingParticleNet

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports a direct search for the Higgs boson decaying to a charm quark-antiquark pair, a process the standard model predicts at a small rate and that has not yet been observed. Using 138 fb$^{-1}$ of 13 TeV proton-proton collisions in the $t\bar{t}H$ production mode, it sets an observed (expected) upper limit of 7.8 (8.7) times the standard model rate for $t\bar{t}H(H\to c\bar{c})$. Combined with the earlier $VH$ search, the observed (expected) interval on the charm Yukawa coupling modifier is $|\kappa_c| < 3.5$ ($|\kappa_c| < 2.7$), which the paper identifies as the most stringent constraint on $\kappa_c$ to date. The same analysis measures $t\bar{t}H(H\to b\bar{b})$ relative to the standard model at $0.91^{+0.26}_{-0.22}$, a 4.4-standard-deviation observation, and validates the search strategy by measuring $t\bar{t}Z$ decays to charm and bottom quark pairs. If correct, this result tightens the experimental picture of how the Higgs boson couples to second-generation quarks and sharpens the target for future searches.

What carries the argument

The argument is carried by a binned profile likelihood fit to event counts in signal and control regions defined by a transformer-based event classifier (ParT) and by two ParticleNet jet-tagging discriminants, $p_{B+C}$ and $p_{B\rm vs C}$, which separate heavy-flavor from light-flavor jets and b jets from c jets. The fit treats the five $t\bar{t}+$jets background components as independently floating normalizations constrained in dedicated control regions, and extracts signal strength modifiers for $t\bar{t}H(H\to c\bar{c})$, $t\bar{t}H(H\to b\bar{b})$, $t\bar{t}Z(Z\to c\bar{c})$, and $t\bar{t}Z(Z\to b\bar{b})$. The $\kappa_c$ interpretation uses the kappa-framework, which expresses the $H\to c\bar{c}$ and $H\to b\bar{b}$ branching fractions as functions of the Yukawa coupling modifiers $\kappa_c$ and $\kappa_b$, and combines the $t\bar{t}H$ likelihood with the previous $VH$ result.

What would settle it

Rerun the profile likelihood with the five $t\bar{t}+$jets background components normalized freely in the signal regions instead of being tied to the control regions; if the resulting 95% interval on $\kappa_c$ no longer respects $|\kappa_c| < 3.5$, the headline bound depends on the control-region transfer assumption rather than being robustly determined by the data.

Watch

Extended reading notes

Core claim

The paper's central claim is that a direct $t\bar{t}H$-associated search for $H\to c\bar{c}$, combined with the earlier $VH$ search, yields the tightest bound yet on the Higgs-charm Yukawa coupling: observed $|\kappa_c| < 3.5$ at 95% confidence, with an expected bound of $|\kappa_c| < 2.7$. In the $t\bar{t}H$ channel alone, no signal is established and the observed (expected) 95% CL upper limit on the $t\bar{t}H(H\to c\bar{c})$ rate is 7.8 (8.7) times the standard model prediction. The companion $H\to b\bar{b}$ measurement returns a signal strength of $0.91^{+0.26}_{-0.22}$, a 4.4-standard-deviation effect, while the two $t\bar{t}Z$ validation modes ($Z\to c\bar{c}$ and $Z\to b\bar{b}$) agree with the SM within one and two standard deviations, respectively. The $\kappa_c$ interval is obtained by profiling $\kappa_c$ and $\kappa_b$ in the kappa-framework and then combining the $t\bar{t}H$ likelihood with the $VH$ result, giving the paper's headline result.

Load-bearing premise

The result assumes that, after calibration, the simulation correctly predicts how often charm-quark jets are tagged and how much background from top-quark pairs with extra jets contaminates each region; if either prediction is wrong, the extracted signal strengths and the quoted $\kappa_c$ limits are biased.

Editorial extensions

If this is right

The observed $|\kappa_c| < 3.5$ is the most stringent experimental limit on the charm Yukawa coupling, improving on the previous $VH$-only bounds reported by both major Higgs experiments.
The $t\bar{t}H(H\to b\bar{b})$ signal strength of $0.91^{+0.26}_{-0.22}$ and its 4.4-standard-deviation significance support the $t\bar{t}H$ event-selection and background-modeling chain used for the charm search.
The $t\bar{t}Z(Z\to c\bar{c})$ and $t\bar{t}Z(Z\to b\bar{b})$ measurements, which agree with the standard model, validate that charm-jet tagging and the background estimation method perform as expected in this final state.
If the true charm Yukawa coupling is close to its standard model value, the expected bound of $|\kappa_c| < 2.7$ implies that substantially more integrated luminosity will be required before a $H\to c\bar{c}$ signal can be established.
The simultaneous measurement of $H\to b\bar{b}$ and $H\to c\bar{c}$ in the same production mode offers a direct experimental handle on the ratio $\kappa_c/\kappa_b$, with production-side uncertainties suppressed.

Reading between the lines

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

Because the leading uncertainty on the $H\to c\bar{c}$ signal strength is statistical (roughly 74%), the same analysis at a higher-luminosity LHC should tighten the $\kappa_c$ bound roughly as the square root of the accumulated luminosity before systematic uncertainties dominate.
The observed limit on $\kappa_c$ being less stringent than expected (3.5 observed versus 2.7 expected) is consistent with a small upward fluctuation in the charm-like signal categories; the current evidence is far from conclusive, but a future run could test whether any hint of an enhanced charm coupling persists.
The combination logic used here could be extended to include gluon-fusion and vector-boson-fusion production of the Higgs boson, which would further constrain $\kappa_c$ if charm-tagging performance at lower Higgs transverse momenta can be controlled.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Bottom line: this is a solid, incremental result from CMS. The genuinely new piece is the ttH-production search for H->cc; previous CMS and ATLAS limits came from VH only. The analysis is standard but well executed: ParticleNet for jet flavor, ParT as event classifier, a simultaneous fit of ttH(H->bb), ttH(H->cc), ttZ(Z->bb), and ttZ(Z->cc), with five floating tt+jets backgrounds. The ttZ validation (1.02 +0.79/-0.84 for Z->cc, 1.47 +0.45/-0.41 for Z->bb) is a genuine check and gives confidence that charm-tagging and background modeling are in the right ballpark.

The headline result, combined observed (expected) |kappa_c| < 3.5 (2.7), is plausible and more stringent than the ATLAS 4.2 and previous CMS 5.5 values. The stress-test note raises an apparent tension: the paper also quotes |kappa_c| < 3.0 for ttH-only when fixing kappa_b to unity. That number is tighter but comes from a more restrictive assumption; the combined bound profiles over kappa_b, so the two are not directly comparable. The loosening from 3.0 to 3.5 in the observed combined limit is likely due to the upward fluctuation in the VH channel (observed limit 14 vs expected 7.6). This does not invalidate the "most stringent" claim, but the paper should say explicitly that 3.0 is conditional on kappa_b = 1 and 3.5 is the profiled result. Without that caveat, a careful reader could fairly call the abstract misleading.

The main limitation is that this is a short proceedings: key systematic details, the likelihood, and the charm-tagging calibration live in CMS-PAS-HIG-24-018. That is normal for conference write-ups, but independent verification has to wait for the PAS or the journal paper. The citation pattern is clean; the one self-reference to the PAS is appropriate.

No serious flaws in the math or logic that I can see. The five float tt+jets normalizations are reasonable and constrained by control regions, and the quoted uncertainties are internally consistent.

Who is this for? Higgs physicists and flavor physicists tracking Yukawa couplings. It is not a methodological breakthrough; it is a careful, incremental measurement. It deserves a serious referee if submitted to a journal; as a proceedings, it is fine once the kappa_b assumption is clarified. I would cite it in any review of Higgs couplings.

Referee Report

1 major / 3 minor

Summary. These conference proceedings report a CMS search for the standard model Higgs boson decay H→cc in association with a top quark-antiquark pair (ttH), using 138 fb^-1 of 13 TeV proton-proton collisions. The analysis measures the ttH(H→bb) rate simultaneously, validates the method by measuring ttZ production, and sets an observed (expected) 95% CL upper limit on mu_ttH(H→cc) of 7.8 (8.7). The result is interpreted in the kappa framework and combined with an earlier VH search, yielding observed (expected) |kappa_c| < 3.5 (< 2.7), which the paper describes as the most stringent constraint on kappa_c to date.

Significance. If the quoted result is correct, it would provide the tightest direct constraint on the Higgs-charm Yukawa coupling, a key test of the standard model's second-generation fermion couplings. The analysis is technically sophisticated, using a ParticleNet-based flavor tagger and a ParT multiclass event classifier, and it simultaneously measures ttH(H→bb) and validates the approach with ttZ. The proceedings is concise and relies on the companion CMS-PAS-HIG-24-018 for systematic and technical details; this is appropriate for a conference summary. The central physics result is plausible and consistent with the numbers shown, but the headline claim of being the 'most stringent constraint' is not fully supported by the numbers as presented.

major comments (1)

[Section 3] The statement 'This represents the most stringent constraint on kappa_c to date' is difficult to reconcile with the same section's earlier quoted result, obtained in the ttH-only analysis with kappa_b fixed to unity, of observed (expected) |kappa_c| < 3.0 (3.3). The combined observed limit is |kappa_c| < 3.5, which is numerically looser than the ttH-only fixed-kappa_b limit. The text does not specify the statistical treatment used for the combined result (e.g., whether kappa_b is profiled rather than fixed) nor why the two numbers are not directly comparable. If the ttH-only result with kappa_b profiled yields an observed upper bound below 3.5, then the combined result is not the most stringent observed constraint. I request that the authors clarify this point, either by providing the profiled ttH-only value or by qualifying the claim to refer to a specific statistical framework (e.g., 'in a combined fit with kappa_b allowed to float'). Without this clarification, the central claim of the paper is ambiguous.

minor comments (3)

[Abstract] The notation 'kappa_c < 3.5' and 'kappa_c < 2.7' should be '|kappa_c| < 3.5' and '|kappa_c| < 2.7' to match the body text and the convention for a two-sided parameter.
[Section 3] The phrase 'observed (expected) 95% CL interval is |kappa_c| < 3.0 (3.3)' is unconventional because an upper limit is a one-sided interval; consider using 'upper limit' for clarity.
[Section 2] In the sentence describing the floating tt+jets normalizations, the list 'tt+c, tt+≥2c, tt+b, tt+≥2b, and tt+light' would be clearer if set off with semicolons or presented as a parenthetical list.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ttH and combined kappa_c limits are derived from a profile likelihood fit to independent data, and the VH combination is a separate prior CMS measurement.

full rationale

The paper's derivation chain is self-contained. The ttH(H->cc) and ttH(H->bb) signal strengths are extracted from a binned profile likelihood fit to 138 fb^-1 of CMS data using ttagging categories and a multiclass ParT classifier; the observed (expected) 95% CL upper limit on mu_ttH(H->cc) of 7.8 (8.7) is an output of that fit, not an input. The combined |kappa_c| < 3.5 (2.7) interval is obtained by statistically combining this ttH search with the previous CMS VH search (Ref. 10), which uses an independent dataset and is not a fitted parameter of the present analysis. References to CMS-PAS-HIG-24-018 (Ref. 13) point to the detailed PAS for efficiencies and figures, serving as provenance rather than as a load-bearing self-justifying premise. No equation in the paper defines a prediction in terms of the quantity it claims to predict, no fitted parameter is renamed as a prediction, and no uniqueness argument is imported from prior work by the same authors. The possible statistical subtlety that the ttH-only observed interval |kappa_c| < 3.0 is quoted with kappa_b fixed to unity while the combined interval profiles kappa_b is a matter of limit-setting convention and comparability, not circularity; it does not make the result equivalent to its inputs by construction.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The analysis is an experimental measurement with no invented entities. The free parameters are nuisance normalizations for the dominant background; the signal strengths are the results of the fit, not ad hoc inputs. The key assumptions are the accuracy of the simulation-based taggers and the kappa framework interpretation.

free parameters (5)

Normalization of tt+c background = not quoted
Floating in the profile likelihood fit, constrained by a dedicated control region.
Normalization of tt+>=2c background = not quoted
Floating in the profile likelihood fit, constrained by a dedicated control region.
Normalization of tt+b background = not quoted
Floating in the profile likelihood fit, constrained by a dedicated control region.
Normalization of tt+>=2b background = not quoted
Floating in the profile likelihood fit, constrained by a dedicated control region.
Normalization of tt+light background = not quoted
Floating in the profile likelihood fit, constrained by a dedicated control region.

assumptions (4)

domain assumption The kappa-framework parameterization (Ref. 17) correctly maps signal strengths to Yukawa coupling modifiers.
Used to interpret the measured limits as constraints on kappa_c and kappa_b. This assumes the framework and SM production cross sections.
domain assumption The ParticleNet and ParT classifiers, trained on simulation, provide accurate jet flavor identification after calibration.
The analysis relies on these algorithms to separate signal from backgrounds; any calibration bias directly affects the result.
domain assumption The tt+jets background shape is correctly modeled by simulation, with only normalizations floating.
The background composition is taken from simulation; the fit only adjusts overall normalizations per component.
domain assumption The integrated luminosity of 138 fb^-1 is accurate.
All yields are scaled to this luminosity; an error would shift the measured signal strengths proportionally.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Search for Higgs boson decay to a charm quark-antiquark pair via $t\bar{tH}$ production." pith.science (2026). https://pith.science/paper/TPR6GRNZ

@misc{pith2026250602163,
  author       = {Pith},
  title        = {Pith review of: Search for Higgs boson decay to a charm quark-antiquark pair via $t\bartH$ production},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TPR6GRNZ}},
  note         = {Machine review of arXiv:2506.02163}
}

abstract

In these proceedings, a search for the standard model Higgs boson decaying to a charm quark-antiquark pair, $H\to c\bar{c}$, produced in association with a top quark-antiquark pair ($t\bar{tH}$) is presented. The search is performed using proton-proton collision data collected by the CMS experiment at $s=13$ TeV, corresponding to an integrated luminosity of 138 fb$^{-1}$. The Higgs boson decay to a bottom quark-antiquark pair is measured simultaneously and the observed $t\bar{tH(H\to bb)}$ event rate relative to the standard model expectation is found to be $0.91^{+0.26}_{-0.22}$. The observed (expected) upper limit at 95\% confidence level (CL) for $t\bar{tH(H\to cc)}$ production is 7.8 (8.7) times the standard model prediction. Combined with a previous search for $H\to c\bar{c}$ via associated production with a W or Z boson, the observed (expected) 95\% CL interval on the Higgs-charm Yukawa coupling modifier, $\kappa_{c}$, is $\kappa_{c} < 3.5$ ($\kappa_{c} < 2.7$).

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

[1]

JINST, 3:S08003, 2008

ATLAS Collaboration. JINST, 3:S08003, 2008

work page 2008
[2]

JINST, 3:S08004, 2008

CMS Collaboration. JINST, 3:S08004, 2008

work page 2008
[3]

ATLAS Collaboration. Phys. Lett. B, 716:1, 2012

work page 2012
[4]

CMS Collaboration. Phys. Lett. B, 716:30, 2012

work page 2012
[5]

CMS Collaboration. Phys. Lett. B, 805:135425, 2020

work page 2020
[6]

Nature, 607:60, 2022

CMS Collaboration. Nature, 607:60, 2022

work page 2022
[7]

Nature, 607(7917):52–59, 2022

ATLAS Collaboration. Nature, 607(7917):52–59, 2022. [Erratum: Nature 612, E24 -4 -3 -2 -1 0 1 2 3 4 b -20 -15 -10 -5 0 5 10 15 20 c 138 fb 1 (13 TeV) CMS Preliminary Observed SM Expected 68% CL 95% CL 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 95% CL upper limit on H cc VH(H cc) ttH(H cc) Combined 138 fb 1 (13 TeV) CMS Preliminary Exp. 7.6 Obs. 14 Exp. 8.7 Obs. ...

work page 2022
[8]

CMS Collaboration. Phys. Rev. Lett., 122:021801, 2019

work page 2019

Show all 17 references

[9]

Measurements of WH and ZH production with Higgs boson decays into bottom quarks and direct constraints on the charm Yukawa coupling in 13 TeV pp collisions with the ATLAS detector

ATLAS Collaboration. Measurements of WH and ZH production with Higgs boson decays into bottom quarks and direct constraints on the charm Yukawa coupling in 13 TeV pp collisions with the ATLAS detector. Submitted to JHEP, 2024

2024
[10]

CMS Collaboration. Phys. Rev. Lett., 131:061801, 2023

2023
[11]

Huilin Qu and Loukas Gouskos. Phys. Rev. D, 101:056019, 2020

2020
[12]

JINST, 15:P12012, 2020

Emil Bols, Jan Kieseler, Mauro Verzetti, Markus Stoye, and Anna Stakia. JINST, 15:P12012, 2020

2020
[13]

Search for Higgs boson decay to a charm quark-antiquark pair via t¯tH production, 2023

CMS Collaboration. Search for Higgs boson decay to a charm quark-antiquark pair via t¯tH production, 2023. CMS-PAS-HIG-24-018 https://cds.cern.ch/record/2929444

2023
[14]

Particle Transformer for jet tagging

Huilin Qu, Congqiao Li, and Sitian Qian. Particle Transformer for jet tagging. In Proceedings of the 39th International Conference on Machine Learning (ICML), Baltimore, USA, 2022. [PMLR 162:18281-18292]

2022
[15]

Inclusive and differential measurement of top quark cross sections in association with a z boson, 2024

CMS Collaboration. Inclusive and differential measurement of top quark cross sections in association with a z boson, 2024. CMS-PAS-TOP-23-004 https://cds.cern.ch/record/ 2893862

2024
[16]

JHEP, 07:163, 2024

ATLAS Collaboration. JHEP, 07:163, 2024

2024
[17]

Handbook of LHC Higgs cross sections: 4

LHC Higgs Cross Section Working Group. Handbook of LHC Higgs cross sections: 4. deciphering the nature of the Higgs sector. CERN Yellow Report, CERN, 2016. CERN- 2017-002

2016

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

JINST, 3:S08003, 2008

ATLAS Collaboration. JINST, 3:S08003, 2008

work page 2008

[2] [2]

JINST, 3:S08004, 2008

CMS Collaboration. JINST, 3:S08004, 2008

work page 2008

[3] [3]

ATLAS Collaboration. Phys. Lett. B, 716:1, 2012

work page 2012

[4] [4]

CMS Collaboration. Phys. Lett. B, 716:30, 2012

work page 2012

[5] [5]

CMS Collaboration. Phys. Lett. B, 805:135425, 2020

work page 2020

[6] [6]

Nature, 607:60, 2022

CMS Collaboration. Nature, 607:60, 2022

work page 2022

[7] [7]

Nature, 607(7917):52–59, 2022

ATLAS Collaboration. Nature, 607(7917):52–59, 2022. [Erratum: Nature 612, E24 -4 -3 -2 -1 0 1 2 3 4 b -20 -15 -10 -5 0 5 10 15 20 c 138 fb 1 (13 TeV) CMS Preliminary Observed SM Expected 68% CL 95% CL 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 95% CL upper limit on H cc VH(H cc) ttH(H cc) Combined 138 fb 1 (13 TeV) CMS Preliminary Exp. 7.6 Obs. 14 Exp. 8.7 Obs. ...

work page 2022

[8] [8]

CMS Collaboration. Phys. Rev. Lett., 122:021801, 2019

work page 2019