Pith. Sign in

REVIEW

Search for a light pseudoscalar Higgs boson in final states with boosted muon pairs and tau lepton pairs in proton-proton collisions at $\sqrt{s}$ = 13 TeV

PaperPDF from arXiv

data files

ReviewThe objections and verdict, as JSON Paper breakdownThe authors' claims, conditions and measurements, as JSON Signed receiptProof this review has not been altered
X Bluesky LinkedIn Reddit Hacker News
CMS Collaboration

High Energy Physics - Experiment · Submitted 11 Sep 2026 · 88 references

Pith's verdict: Solid CMS search with a real blind spot in the τhτh signal efficiency.

T0 · AI review

How this review was made: An AI referee (deepseek/deepseek-v4-flash) read the paper on September 14, 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.

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

How Pith checks papers →

· 1 major objection · 4 minor · Ask about this paper · The compiled record · Read the full review →
Challenge this review Comment (0) Read on arXiv ↗
Cite ▾
arXiv ID2609.12683 (version 1)
Pith IDLRCAPEYM, Pith's permanent record of this paper and its review
CitationCMS Collaboration. "Search for a light pseudoscalar Higgs boson in final states with boosted muon pairs and tau lepton pairs in proton-proton collisions at $\sqrt{s}$ = 13 TeV." arXiv:2609.12683 (2026). https://pith.science/paper/LRCAPEYM
Cite this reviewBibTeX for Pith's review of this paper
Claim paper

Why Pith says this

Highlight any text to comment on it.

Referee report

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

CMS sets stringent limits on Higgs decays to a light pseudoscalar in the μμττ final state with boosted topologies.

The reading

This paper searches for the process H → aa → μμττ, where H is the 125 GeV Higgs boson or a heavier scalar (up to 1 TeV) and a is a light pseudoscalar (3.6 to 50 GeV). Because the large mass difference forces the tau leptons from a → ττ to be highly collimated, standard tau reconstruction fails. The analysis develops dedicated boosted-tau techniques, including a deep neural network tagger for hadronic ditau pairs, and combines four tau decay channels. Using the full CMS Run 2 dataset (138 fb⁻¹ at 13 TeV), no significant excess over Standard Model background is observed. The result sets the strongest model-independent limits to date on σ_H × B(H→aa→μμττ)/σ_SM at low pseudoscalar masses and, for the first time, extends these limits to heavy scalar masses up to 1 TeV. If the paper is right, any sizable branching fraction for H → aa with subsequent decays to μμττ is excluded, constraining models such as the NMSSM and 2HDM+S in previously unexplored low-mass regions.

What carries the argument

The central object enabling the search is a set of dedicated boosted-ditau reconstruction techniques. For the τ_μ τ_h and τ_e τ_h channels, a modified hadron-plus-strips (HPS) reconstruction removes the overlapping lepton from the jet seed and relaxes the isolation cone. For the τ_h τ_h channel, the DEEPDITAU deep neural network classifies a reconstructed AK4 jet as originating from a collimated hadronic ditau pair, a light jet, or a b jet, achieving 70% tagging efficiency with a ~1% mistag rate. These techniques together recover sensitivity to the 94% of a→ττ decays in the boosted regime where standard tau identification fails.

Extended reading notes

Core claim

Using 138 fb⁻¹ of proton-proton collisions at √s = 13 TeV, the CMS Collaboration finds no evidence for the decay of a scalar boson (either the 125 GeV Higgs or a heavier neutral scalar) into two light pseudoscalar bosons that subsequently decay to a muon pair and a tau lepton pair. For the 125 GeV Higgs boson, 95% confidence-level upper limits on σ_H × B(H→aa→μμττ)/σ_SM range from 5×10⁻⁵ to 2.7×10⁻⁴ for pseudoscalar masses between 3.6 and 21 GeV. For a hypothetical scalar with mass 1 TeV, limits range from 3.5×10⁻⁴ to 1.5×10⁻³ for pseudoscalar masses between 3.6 and 50 GeV.

Load-bearing premise

The DEEPDITAU tagger, which identifies collimated τ_h τ_h pairs in the high-background τ_h τ_h channel, is trained only on simulated samples, and its mistag rate (approximately 1%) and efficiency (70%) are not directly validated with collision data — any data-simulation discrepancy in the tagger performance would directly affect the signal sensitivity and background estimate in that channel.

Editorial extensions

If this is right

  • The limits on σ_H×B(H→aa→μμττ)/σ_SM for m_H = 125 GeV directly constrain the branching fraction of the Higgs boson to a pair of light pseudoscalars in any model where the dominant a decays are to leptons, including Type-III 2HDM+S and NMSSM scenarios.
  • For the first time at the LHC, upper limits on H → aa in the μμττ final state are set for heavy scalar masses up to 1 TeV, extending the reach to previously unprobed heavy-resonance parameter space.
  • The derived 2HDM+S exclusion contours (Types I–IV) provide the most stringent limits on B(H→aa) for m_a between 3.6 and 4.2 GeV, and complement existing constraints from the bbbb, bbττ, bbμμ, and 4τ final states.
  • The data-driven background estimation using the tight-to-loose ratio in a Z→μμ+jet control region provides a method that can be adapted to other searches for boosted di-tau resonances at the LHC.

Reading between the lines

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

  • The DEEPDITAU tagger, being trained only on simulation and not validated with data, carries an unquantified systematic uncertainty; a dedicated data validation using e.g. γ → ττ or Z → ττ events with boosted topologies would strengthen the confidence in the τ_h τ_h channel results.
  • The combination of four tau decay channels (τ_μ τ_e, τ_μ τ_h, τ_e τ_h, τ_h τ_h) achieves 94% coverage of all a→ττ decays; future searches could aim to include the remaining 6% (fully leptonic tau pairs in boosted topologies) to close the coverage entirely.
  • The reported limits for the 125 GeV Higgs boson at low m_a (3.6–21 GeV) are now at the level of a few times 10⁻⁵ of the Standard Model cross section; this begins to probe the natural parameter space of some NMSSM benchmarks where B(H→aa) can be as small as 1–5%.
Original abstract

A search is presented for a light pseudoscalar Higgs boson, a, using LHC proton-proton data at $\sqrt{s}$ = 13 TeV collected by the CMS experiment from 2016 to 2018, corresponding to an integrated luminosity of 138 fb$^{-1}$. The analysis targets the process H $\to$ aa $\to \mu\mu\tau\tau$, where the scalar H can be either the 125 GeV Higgs boson or a heavier state. Because of the large mass difference between H and a, the two tau leptons are highly Lorentz boosted and collimated, often failing standard CMS tau lepton reconstruction. To address this, dedicated ditau reconstruction techniques, including a deep neural network, are developed to improve the efficiency for events in which at least one tau lepton decays to hadrons. No significant excess over the standard model (SM) background is observed. Model-independent upper limits at the 95% confidence level are set on the product of the Higgs boson production cross section and branching fraction, divided by the SM cross section. For a Higgs boson mass ($m_\mathrm{H}$) equal to 125 GeV, these limits range from 5$\times$10$^{-5}$ to 2.7$\times$10$^{-4}$ for pseudoscalar masses between 3.6 and 21 GeV. For the first time using LHC data, upper limits are also obtained in final states with tau leptons for a hypothetical scalar boson with a mass up to 1 TeV. For $m_\mathrm{H}$ = 1 TeV, the limits range from 3.5$\times$10$^{-4}$ to 1.5$\times$10$^{-3}$ for pseudoscalar masses between 3.6 and 50 GeV. In addition, limits are set for SM extensions involving two-Higgs-doublet plus singlet models for $m_\mathrm{H}$ = 125 GeV, extending previous LHC results.

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, 8,044 input tokens.

What the abstract itself states, as probabilities: claims first 0.87 · states limitation 0.66 · human participants 0.01 · controlled comparison 0.50 · randomized assignment 0.01 · states quantitative result 0.99 · compares with prior results 0.77 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; .

A search is presented for a light pseudoscalar Higgs boson, a, using LHC proton-proton data at $\sqrt{s}$ = 13 TeV collected by the CMS experiment from 2016 to 2018, corresponding to an integrated luminosity of 138 fb$^{-1}$. The analysis targets the process H $\to$ aa $\to \mu\mu\tau\tau$, where the scalar H can be either the 125 GeV Higgs boson or a heavier state. Because of the large mass difference between H and a, the two tau leptons are highly Lorentz boosted and collimated, often failing standard CMS tau lepton reconstruction. To address this, dedicated ditau reconstruction techniques, including a deep neural network, are developed to improve the efficiency for events in which at least one tau lepton decays to hadrons. No significant excess over the standard model (SM) background is observed. Model-independent upper limits at the 95% confidence level are set on the product of the Higgs boson production cross section and branching fraction, divided by the SM cross section. For a Higgs boson mass ($m_\mathrm{H}$) equal to 125 GeV, these limits range from 5$\times$10$^{-5}$ to 2.7$\times$10$^{-4}$ for pseudoscalar masses between 3.6 and 21 GeV. For the first time using LHC data, upper limits are also obtained in final states with tau leptons for a hypothetical scalar boson with a mass up to 1 TeV. For $m_\mathrm{H}$ = 1 TeV, the limits range from 3.5$\times$10$^{-4}$ to 1.5$\times$10$^{-3}$ for pseudoscalar masses between 3.6 and 50 GeV. In addition, limits are set for SM extensions involving two-Higgs-doublet plus singlet models for $m_\mathrm{H}$ = 125 GeV, extending previous LHC results.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#sentencerolethe authors' own workstates the main conclusionstrength
s0 A search is presented for a light pseudoscalar Higgs boson, a, using LHC proton-proton data at $\sqrt{s}$ = 13 TeV collected by the CMS experiment from 2016 to 2018, corresponding to an integrated luminosity of 138 fb$^{-1}$. method 0.50 0.89 0.02 asserted 0.99
s1 The analysis targets the process H $\to$ aa $\to \mu\mu\tau\tau$, where the scalar H can be either the 125 GeV Higgs boson or a heavier state. contribution 0.88 0.44 0.02 asserted 0.92
s2 Because of the large mass difference between H and a, the two tau leptons are highly Lorentz boosted and collimated, often failing standard CMS tau lepton reconstruction. background 0.86 0.35 0.02 asserted 0.97
s3 To address this, dedicated ditau reconstruction techniques, including a deep neural network, are developed to improve the efficiency for events in which at least one tau lepton decays to hadrons. method 0.92 0.95 0.02 asserted 1.00
s4 No significant excess over the standard model (SM) background is observed. result 1.00 0.95 0.89 asserted 1.00
s5 Model-independent upper limits at the 95% confidence level are set on the product of the Higgs boson production cross section and branching fraction, divided by the SM cross section. result 0.97 0.96 0.13 asserted 1.00
s6 For a Higgs boson mass ($m_\mathrm{H}$) equal to 125 GeV, these limits range from 5$\times$10$^{-5}$ to 2.7$\times$10$^{-4}$ for pseudoscalar masses between 3.6 and 21 GeV. result 1.00 0.96 0.14 asserted 1.00
s7 For the first time using LHC data, upper limits are also obtained in final states with tau leptons for a hypothetical scalar boson with a mass up to 1 TeV. contribution 0.89 0.95 0.05 asserted 1.00
s8 For $m_\mathrm{H}$ = 1 TeV, the limits range from 3.5$\times$10$^{-4}$ to 1.5$\times$10$^{-3}$ for pseudoscalar masses between 3.6 and 50 GeV. result 1.00 0.95 0.08 asserted 1.00
s9 In addition, limits are set for SM extensions involving two-Higgs-doublet plus singlet models for $m_\mathrm{H}$ = 125 GeV, extending previous LHC results. contribution 0.48 0.92 0.04 asserted 1.00

What kind of paper the classifier reports this is: empirical study 1.00 · evidence from analysis of existing data 0.72.

What was examined, and what was not

sources read
  • arXiv listing abstract v1
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. Feynman diagram of a Higgs boson produced via gluon fusion and decaying into two
Figure 2. The ττ reconstruction efficiency of the standard τ HPS algorithm (open markers) and the modified τµ τh (left) or τe τh (right) HPS algorithm (solid markers) as a function of ma for mH = 125, 250, 500, 750, and 1000 GeV. Because of the requirement ∆R(τℓ , τh ) < 0.8, the efficiency for mH = 125 GeV decreases above ma = 12 GeV in the τµ τh channel and above ma = 14 GeV in the τe τh channel, corresponding to the region in which the ττ topology becomes increasingly resolved. In the right plot, the dip in the standard HPS efficiency for the low ma region is due to two competing eff
Figure 3. Receiver operating characteristic curve for the D
Figure 4. Schematic of the analysis strategy. In the Z boson mass range, the tight-to-loose ratio
Figure 5. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 6. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 7. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 8. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 9. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 10. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 11. Projections of the post-fit two-dimensional background pdfs (blue) and the observed
Figure 12. Expected and observed limits on σH × B(H → aa → µµττ)/σSM, combining the τµ τe , τµ τh , τe τh , and τh τh final states, for mH = 125, 250, 500, 750, and 1000 GeV. The vertical dotted lines indicate the boundaries of the mµµ regions used in the background fit, as described in Section 7
Figure 13. Expected and observed limits on σH × B(H → aa → µµττ)/σSM, combining the final states for different Higgs boson mass hypotheses, for the pseudoscalar masses of 10 GeV (left) and 15 GeV (right). The straight lines between points are there to guide the eye. the high-rate but high-background τh τh final state. Data collected by the CMS experiment at √ s = 13 TeV, corresponding to an integrated luminosity of 138 fb−1 , are examined. No signifi￾cant excess over SM processes is observed. Model-independent upper limits at the 95% confi￾dence level are set on the product of the Higgs boson…
Figure 14. Observed and expected limits in 2HDM+S models of Types I through IV. The con

Figures reproduced from arXiv:2609.12683 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.