Improved search for psi(3770) to gamma eta c(1S, 2S) radiative transitions
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: High Energy Physics: Experiment. The day in high-energy experiment: detectors, searches and limits.
Kai: I'm Kai.
Mira: And I'm Mira, here with Lev.
Kai: We're looking at the paper: "Improved search for psi(3770) to gamma eta c(1S, 2S) radiative transitions".
Mira: Here's the gist of the paper. Improved search for psi(3770) → γηc(1S, 2S) radiative transitions is presented using an improved data sample to set upper limits on branching fractions and partial decay widths.
Kai: Experimental Setup and Data Utilization The search was conducted using an integrated luminosity of 20.3 fb−1 of e+e− annihilation data collected at a center-of-mass energy of 3.773 GeV with the BESIII detector operating at the BEPCII collider.
Mira: The signal decay channels ψ(3770) → γηc(1S, 2S) are reconstructed via ηc(1S, 2S) → K0SK±π∓.
Kai: This reconstruction benefits from a large branching fraction, high reconstruction efficiency, and a low combinatorial background level.
Paper summary: Kai: So, wrapping up this discussion on "Improved search for psi(three thousand seven hundred seventy) to gamma eta c(1S, 2S) radiative transitions," what does it really mean for the field <ref:2609.33606#pg1>? The main thing is that they've managed to reduce those upper limits on the partial decay widths by about thirty and fifty percent compared to previous results.
Mira: That reduction is significant because it brings them closer to some of the theoretical predictions, specifically for psi three seven seventy zero into gamma eta c one S. The upper limit on that partial decay width is now within a one sigma range of calculations from lattice quantum chromodynamics and other models.
Lev: But they also found that the upper limit on the second transition, psi three seven seventy zero into gamma eta c two S, is about ten times larger than what some of those lattice calculations predict. It's significantly larger than what you get when you use the relativistic BS equation method with its large relativistic corrections.
Kai: So it’s a mixed message then. We have better constraints on one channel, but the other one is still quite open, or maybe that theoretical prediction for the second transition is just not quite matching what we see here.
Mira: Exactly. The paper shows that while the search method itself has improved, there's still a gap between what we measure and what some of our theoretical methods predict for those specific decay widths.
Lev: The limitation they state is that their uncertainty on the product branching fractions is quite large, with multiplicative uncertainties reaching up to sixty three point five percent. That means when you interpret those final numbers, you need to be very careful about how much weight you give them.
Kai: So, the paper is a solid step forward in the experimental side of things because they've pushed those limits down by a lot and provided concrete numbers for these specific transitions.
Mira: It’s not a complete solution, though. The paper clearly shows that the second transition limit is still quite high compared to what some established theoretical frameworks suggest for the eta c two S state.
Lev: The implication is that either we need better experimental sensitivity in those regions, or we need to refine our theoretical models for those specific decay widths.
Kai: So, in short, it’s an improvement on the search, but it highlights where the current theoretical understanding might still be incomplete regarding these radiative decays.
Conclusion: Kai: So, this paper is about improving how we look for two specific radiative transitions from the psi three seven seventy zero state to eta c one S and eta c two S states.
Mira: Yeah, they used a better data sample than before to set upper limits on those branching fractions and partial decay widths.
Lev: What did they actually do in terms of the experiment? Were there new detector settings or something?
Kai: They used twenty point three femtobarns of e plus e minus annihilation data from the BESIII detector at three point seven seven three giga electron volt with a center of mass energy.
Mira: And they reconstructed these decays by looking for eta c one S and eta c two S states decaying into kaon, sigma, and pion.
Lev: The reconstruction method uses charged tracks detected in the mdc to be within the polar angle range of less than zero point nine three cosine theta.
Kai: They modeled the signal using a probability density function that includes factors like e seven gamma times big omega of m times f d of e gamma times epsilon of m tensor G delta m sigma.
Mira: The background estimation used tools called topo ana to look at inclusive monte carlo samples, and they found four main sources for the background.
Lev: Those include psi three seven seventy zero decaying into d zero bar zero and e plus e minus annihilation into pi naught kaon sigma k plus minus pion minus.
Kai: They also accounted for the radiative tail of the psi three six eight six resonance, which is going to produce some peaking backgrounds in the signal regions.
Mira: The unbinned maximum likelihood fits gave statistical significances of one point one sigma for eta c one S and one point zero sigma for eta c two S.
Lev: Since they didn't see a significant signal for either state, they calculated ninety percent confidence level upper limits on the number of signal events.
Kai: Specifically, the upper limit on the product branching fractions is less than five point zero times ten to the minus six at ninety percent confidence level.
Mira: That means that for both channels, the upper limit on those partial decay widths are less than five point five kilo electron volt and twenty nine point four kilo electron volt respectively.
Lev: The upper limit on eta c one S is within one sigma of some calculations from lattice quantum chromodynamics and other models, but it's larger than what you get from the relativistic bs equation method that has large relativistic corrections.
Kai: And for eta c two S, the upper limit is about ten times larger than those lattice predictions and way bigger than what the relativistic bs equation method gives.
Mira: The total multiplicative uncertainty for those branching fractions is about thirteen point three percent and sixty three point five percent, so you have to be careful with how much weight you give those limits.
Lev: The main conclusion is that the search for these radiative transitions has improved over previous measurements because the upper limits on the partial decay widths are reduced by about thirty and fifty percent.
Kai: So, we have tighter constraints now on these specific decays, bringing them closer to what some of our theoretical calculations predict for eta c one S.
Mira: But as Lev mentioned, the limit on eta c two S is still quite high compared to those lattice calculations and other methods we've used.
Lev: It seems like either we need better experimental sensitivity in those regions or we need to refine the theoretical models for these specific decay widths.
M. Ablikim, M. N. Achasov, P. Adlarson, X. C. Ai, C. S. Akondi A B, R. Aliberti A C, A., Q., α An, ε Anderson M S, Ι Bai Y O Bakina H R Bao X L Bao M Barbagiovanni V Batozskaya K Begzsuren N Berger M Berlowski M B Bertani D Bettoni F Bianchi E Bianco A Bortone I Boyko R A Briere A Brueggemann D Cabiati H Cai M Cai X Cai G F Cao N Cao S Cetin X Y Chai J F Chang T T Chang G R Che Y Z Che C H Chen Chao Chen G Chen H S Chen H Y Chen M L Chen S J Chen S M Chen T. , what_researchers_did, The researchers performed an improved search for the radiative transitions of the psi(3770) resonance to γc(1S, 2S) using data from the BESIII detector., supporting_evidence, No significant signal was observed with 90 percent confidence level upper limits on the product branching fractions set to be 5.0 x 10^-6 for the γc(1S) transition and 3.7 x 10^-6 for the →㬼(2S) transition., surprising_detail, The improved measurements using a seven times larger integrated luminosity lowered the upper limits by approximately a factor of three and two for the γc(1S) and εc(2S) transitions, respectively., why_it_matters, This work provides improved constraints on theoretical predictions for charmonium radiative decays to test nonperturbative mechanisms., limitations, The study relies on a data sample with an integrated luminosity of 20.3 fb-1 and assumes the use of a specific reconstruction method for γc(1S, 2S)., confidence_or_uncertainty, medium
Institute of High Energy Physics · Beihang University · Bochum Ruhr-University · Budker Institute of Nuclear Physics SB RAS · Carnegie Mellon University · Central China Normal University · Central South University · Chengdu University of Technology · Johannes Gutenberg University of Mainz
hep-ex
Submitted: 2026-09-27
Updated: 2026-09-27
DOI: 10.1103/ml3k-2pss
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 84/100
The gist: Improved search for psi(3770) → γηc(1S, 2S) radiative transitions is presented using an improved data sample to set upper limits on branching fractions and partial decay widths.
Key concepts
- psi(3770)
- This is a specific particle studied in particle physics experiments, often produced in electron-positron annihilation. In this context, it's the initial state resonance being searched for transitions from it to other particles.
- gamma eta c(1S, 2S)
- This refers to a specific decay channel where the psi(3770) particle decays into a photon (gamma) and an excited state of the charmed meson, denoted as eta c(1S or 2S). This is the signal being looked for.
- Branching Fraction
- The branching fraction represents the probability that a specific particle decay will occur. The paper sets upper limits on this probability, meaning they are constraining how often psi(3770) decays into gamma eta c(1S or 2S) compared to other possible decays.
- Partial Decay Width
- The partial decay width is a measure of the probability of a particle decaying into a specific final state. The study calculated upper limits on these widths, which are related to the branching fractions and provide constraints on the strength of this particular decay process.
Terminology
Summary
Improved search for psi(3770) → γηc(1S, 2S) radiative transitions is presented using an improved data sample to set upper limits on branching fractions and partial decay widths. p. 3 3 fb−1
of e
+e
− annihilation data collected at a centerof-mass energy of 3. 773 GeV with the BESIII detector operating at the BEPCII collider, an improved search for the radiative transitions ψ(3770) → γηc(1S, 2S) is performed using the hadronic
decays ηc(1S, 2S) → K0
SK±π
∓.Where the paper says it
Experimental Setup and Data Utilization
The search was conducted using an integrated luminosity of 20.3 fb−1 of e+e− annihilation data collected at a center-of-mass energy of 3. p. 3 China
88 Yantai University, Yantai 264005, People’s Republic of China
89 Yunnan University, Kunming 650500, People’s Republic of China
90 Zhejiang University, Hangzhou 310027, People’s Republic of China
91 Zhengzhou University, Zhengzhou 450001, People’s Republic of China
† Deceased
a Also at the Moscow Institute of Physics and Technology, Moscow 141700, Russia
b Also at the Functional Electronics Laboratory, Tomsk State University, Tomsk, 634050, Russia
c Also at the Novosibirsk State University, Novosibirsk, 630090, Russia
d Also at the NRC "Kurchatov Institute", PNPI, 188300, Gatchina, Russia
e Also at Goethe University Frankfurt, 60323 Frankfurt am Main, Germany
f Also at Key Laboratory for Particle Physics, Astrophysics and Cosmology, Ministry of Education; Shanghai Key Laboratory
for Particle Physics and Cosmology; Institute of Nuclear and Particle Physics, Shanghai 200240, People’s Republic of China
g Also at Key Laboratory of Nuclear Physics and Ion-beam Application (MOE) and Institute of Modern Physics, Fudan
University, Shanghai 200443, People’s Republic of China
h Also at State Key Laboratory of Nuclear Physics and Technology, Peking University, Beijing 100871, People’s Republic of
China
i Also at School of Physics and Electronics, Hunan University, Changsha 410082, China
j Also at Guangdong Provincial Key Laboratory of Nuclear Science, Institute of Quantum Matter, South China Normal
University, Guangzhou 510006, China
k Also at MOE Frontiers Science Center for Rare Isotopes, Lanzhou University, Lanzhou 730000, People’s Republic of China
l Also at Lanzhou Center for Theoretical Physics, Lanzhou University, Lanzhou 730000, People’s Republic of China
m Also at Ecole Polytechnique Federale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland
n Also at Helmholtz Institute Mainz, Staudinger Weg 18, D-55099 Mainz, Germany
o Also at Hangzhou Institute for Advanced Study, University of Chinese Academy of Sciences, Hangzhou 310024, China
p Also at Applied Nuclear Technology in Geosciences Key Laboratory of Sichuan Province, Chengdu University of Technology,
Chengdu 610059, People’s Republic of China
Based on an integrated luminosity of 20. 3 fb−1
of e
+e
− annihilation data collected at a centerof-mass energy of 3. 3 fb−1
of e
+e
− annihilation data collected at a centerof-mass energy of 3. 773 GeV with the BESIII detector operating at the BEPCII collider, an improved search for the radiative transitions ψ(3770) → γηc(1S, 2S) is performed using the hadronic
decays ηc(1S, 2S) → K0
SK±π
∓. The ηc(1S, 2S) are reconstructed via ηc(1S, 2S) → K0
SK±π
∓, which benefits from a large branching fraction, high reconstruction efficiency, and a low combinatorial background level. Other prominent decay channels of the ηc(1S, 2S) either
have much lower reconstruction efficiencies or suffer from
high background levels. In this article, improved measurements of the
ψ(3770) → γηc(1S, 2S) transition rates are presented
using a ψ(3770) resonance data sample with an integrated luminosity of 20. The ηc(1S, 2S) are reconstructed via ηc(1S, 2S) → K0
SK±π
∓, which benefits from a large branching fraction, high reconstruction efficiency, and a low combinatorial background level. The
ψ(3770) → γηc(1S, 2S) decays are denoted as the ηc(1S)
and ηc(2S) channels, respectively. EVENT SELECTION
Charged tracks detected in the MDC are required to
be within the polar angle (θ) range | cos θ| < 0.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Signal Modeling and Background Estimation
The signal probability density function for the ηc(1S, 2S) state is modeled as (E7γ × BW(m) × fd(Eγ) × ϵ(m)) ⊗ G(δm, σ). The background estimation involves analyzing inclusive MC samples using the TopoAna tool. p. 5 DATA ANALYSIS
A. Background estimation
Analysis of the inclusive MC samples using the
TopoAna tool [42] indicates that the dominant background contributions come from four sources: (1)
ψ(3770) → D0D¯ 0
, (D0 → π
+K−, D¯ 0 → π
0K0
S
) + c. Background estimation
Analysis of the inclusive MC samples using the
TopoAna tool [42] indicates that the dominant background contributions come from four sources: (1)
ψ(3770) → D0D¯ 0
, (D0 → π
+K−, D¯ 0 → π
0K0
S
) + c. ;
(2) e
+e
− → π
0K0
SK±π
∓ with the π
0 decaying into a
γγ pair; (3) e
+e
− → K0
SK±π
∓ with a fake photon, ISR
photon, or FSR photon; (4) e
+e
− → γISRψ(3686) →
γISRγηc(1S, 2S). 2 and its shape is modeled
using the distribution derived from MC simulation. Background events from the radiative tail of the
ψ(3686) resonance, i. Background events from the radiative tail of the
ψ(3686) resonance, i. , e
+e
− → γISRψ(3686) →
γISRγηc(1S, 2S), contribute to the peaking backgrounds
within the ηc(1S, 2S) signal regions.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Statistical Significance and Limits
The unbinned maximum-likelihood fits resulted in statistical significances of 1.1σ and 1.0σ for the ηc(1S) and ηc(2S) signals, respectively. p. 7 The resulting statistical significances
of the ηc(1S, 2S) signals are 1. 0σ, respectively. 0σ, respectively. Since no significant signal is observed for either the
ηc(1S) or ηc(2S), 90% C. Since no significant signal is observed for either the
ηc(1S) or ηc(2S), 90% C. upper limits on the number
of signal events are computed, NC. upper limits on the number of signal events, including systematic uncertainties (discussed below), are NC. (90%) <
139 for the ηc(1S) channel and NC. (90%) <
139 for the ηc(1S) channel and NC. (90%) < 156 for the
ηc(2S) channel. UPPER LIMITS
The upper limits on the product branching fractions
are calculated as
B1 × B2 <
NC. ϵ · L · σ
0
ψ(3770) ·
1
|1−Π|
2 · (1 + σ) · B3
, (9)
where B1 and B2 represent B(ψ(3770) → γηc(1S, 2S))
and B(ηc(1S, 2S) → K0
SK±π
∓), respectively, NC. 3 fb−1
of e
+e
− annihilation data collected at a centerof-mass energy of 3. 773 GeV with the BESIII detector operating at the BEPCII collider, an improved search for the radiative transitions ψ(3770) → γηc(1S, 2S) is performed using the hadronic
decays ηc(1S, 2S) → K0
SK±π
∓.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Comparison with Theoretical Predictions
The upper limits on the partial decay widths at 90% C. p. 3 7 × 10−6
for the ηc(2S) transition. The 90% confidence level upper limits on the partial decay widths are also reported to be Γ(ψ(3770) → γηc(1S)) < 5. 14 for the ηc(2S) channel. Fit method
The yields of ψ(3770) → γηc(1S, 2S) → γK0
SK±π
∓
are determined via an unbinned maximum-likelihood fit
on the M(K0
SK±π
∓) spectrum from data. The 90% confidence level upper limits on the partial decay widths are also reported to be Γ(ψ(3770) → γηc(1S)) < 5. 5 keV and
Γ(ψ(3770) → γηc(2S)) < 29. 5 keV
and Γ(ψ(3770) → γηc(2S)) < 29. 4 keV based on the
values of B(ηc(1S, 2S) → K0
SK±π
∓) [55–57]. Table II compares our results with the previous
BESIII measurements [24] and various theoretical predictions [20, 21, 23]. The upper limit on Γ(ψ(3770) →
γηc(1S)) is within the 1σ range of the IML [20] and lattice QCD [21] calculations, and is larger than the prediction obtained using the relativistic BS equation method
that incorporates large relativistic corrections [23]. Table II compares our results with the previous
BESIII measurements [24] and various theoretical predictions [20, 21, 23]. The upper limit on Γ(ψ(3770) →
γηc(1S)) is within the 1σ range of the IML [20] and lattice QCD [21] calculations, and is larger than the prediction obtained using the relativistic BS equation method
that incorporates large relativistic corrections [23].Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Systematic Uncertainties
The total multiplicative uncertainty for the upper limits on B(ψ(3770) → γηc(1S, 2S)) is 13.3% and 63.5%. Additive systematic uncertainties are evaluated by varying the fit model and ηc(1S, 2S) parameters. p. 8 The observed difference is taken as the related uncertainty. The additive systematic uncertainties are evaluated by
varying the fit model and ηc(1S, 2S) parameters: (1) the
uncertainty related to the ηc(1S, 2S) resonance parameters is estimated by varying the parameters by one standard deviation from the PDG values [9]; (2) the uncertainty related to the photon energy dependence is estimated by replacing E7
γ with E5
γ
[23]; (3) the uncertainty related to the damping factor is estimated by removing it from the fit; (4) the uncertainty related to the
Gaussian resolution function is estimated by varying the
parameters by one standard deviation; (5) the uncertainties related to the π
0K0
SK±π
∓, K0
SK±π
∓(γISR/γFSR),
and γISRψ(3686) background yields are estimated by
changing their magnitudes by one standard deviation;
(6) the uncertainty related to the shape of the e
+e
− →
K0
SK±π
∓(γISR/γFSR) background is estimated by varying the smooth parameters of the MC simulated shape;
(7) the uncertainties related to the shapes of the
e
+e
− → γISRψ(3686) → γISRγηc(1S, 2S), and e
+e
− →
γISRJ/ψ/ψ(3686) → γISRK0
SK±π
∓ backgrounds are estimated by replacing the MC simulated shape with a
Breit-Wigner function convolved with a single Gaussian
function with floating parameters; (8) the uncertainty related to the flat background is estimated by replacing the
second-order polynomial with a third-order polynomial. The other sources of systematic uncertainties are multiplicative,
and are listed in Table I. The multiplicative terms include tracking efficiency, K0
S
reconstruction, photon reconstruction, kinematic fit, integrated luminosity of the
ψ(3770) data sample, cross section of e
+e
− → ψ(3770),
D0 veto, branching fractions, and MC statistics.Where the paper says it
Where the paper says it
Conclusion
The search for radiative transitions is improved over previous measurements. The upper limits on the partial decay widths at 90% C.L. are reduced to approximately 30% and 54% of the previous results, respectively. p. 9 Our result is compatible with the previous BESIII measurements [24]. Using a data sample seven times larger than
that used for the previous measurements, the upper limits at 90% confidence level on Γ(ψ(3770) → γηc(1S))
and Γ(ψ(3770) → γηc(2S)) are reduced to approximately 30% and 54% of the previous results, respectively. The upper limit on Γ(ψ(3770) →
γηc(1S)) is within the 1σ range of the IML [20] and lattice QCD [21] calculations, and is larger than the prediction obtained using the relativistic BS equation method
that incorporates large relativistic corrections [23]. For
Γ(ψ(3770) → γηc(2S)), due to the limited statistics of
the data sample and the large uncertainty in B(ηc(2S) →
K0
SK±π
∓), the upper limit is approximately one order
of magnitude larger than the IML prediction and significantly larger than the prediction obtained using the
relativistic BS equation method that incorporates large
relativistic corrections. Our result is compatible with the previous BESIII measurements [24]. Using a data sample seven times larger than
that used for the previous measurements, the upper limits at 90% confidence level on Γ(ψ(3770) → γηc(1S))
and Γ(ψ(3770) → γηc(2S)) are reduced to approximately 30% and 54% of the previous results, respectively.Where the paper says it
Where the paper says it
Where the paper says it
TABLE I. Relative multiplicative systematic uncertainties (%) on the measurements of the upper limits on B(ψ(3770) → γηc(1S, 2S)) at 90% C. p. 9 (90%) < 156 for the ηc(2S) channel. Relative multiplicative systematic uncertainties
(%) on the measurements of the upper limits on B(ψ(3770) →
γηc(1S, 2S)) at 90% C. Relative multiplicative systematic uncertainties
(%) on the measurements of the upper limits on B(ψ(3770) →
γηc(1S, 2S)) at 90% C. The multiplicative uncertainties are
assumed to be independent and are added in quadrature to
obtain the total multiplicative uncertainty.Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved signal search methodology for charmonium radiative transitions by implementing an unbinned maximum-likelihood fit using a signal probability density function that incorporates energy dependence via a damping factor fd(Egamma) derived from KEDR experiment data. This allows the system to precisely constrain partial decay widths, specifically Γ(psi(3770) → γηc(1S)) and Γ(psi(3770) → γηc(2S)), achieving reduced uncertainties compared to previous measurements.
-
Enhanced background modeling capabilities by utilizing data-driven methods for estimating backgrounds from e+e- -> π0K0SK±π∓ with a missing soft photon, using a PHSP MC sample to determine the relative efficiency between the signal and background selections. This allows the system to accurately subtract complex contamination sources like e+e- -> π0K0SK±π∓ and those involving ISR/FSR photons.
-
Integration of systematic uncertainty handling via a convolved likelihood method where the likelihood is convolved with a Gaussian function whose width equals the total multiplicative uncertainty. This allows the system to propagate multiplicative uncertainties from tracking efficiency, K0S reconstruction, photon reconstruction, and integrated luminosity directly into the final upper limit calculations (NC.L.(90%)).
-
Refined constraints on theoretical predictions by allowing for parameter variation in fit models and resonance parameters when calculating systematic uncertainties. The system can test theoretical predictions by varying the damping factor fd(Egamma) or replacing E7g with E5g to quantify how sensitive the results are to model assumptions regarding multipole contributions.
-
Direct comparison of experimental limits with various theoretical frameworks, such as Intermediate Meson Loops (IML) and Lattice Quantum Chromodynamics (LQCD). The system can evaluate whether the obtained upper limits on partial decay widths are compatible with IML predictions or relativistic BS equation method calculations, providing a direct test of nonperturbative mechanisms in charmonium decays.
Abstract
Based on an integrated luminosity of 20.3 fb-1 of e+e- annihilation data collected at a center-of-mass energy of 3.773 with the BESIII detector operating at the BEPCII collider, an improved search for the radiative transitions ψ(3770) to γη c(1S, 2S) is performed using the hadronic decays η c(1S, 2S) to K 0 S K plus or minus π minus or plus. No significant signal is observed. The corresponding 90 % confidence level upper limits on the product branching fractions are set to be 5.0 times 10-6 for the η c(1S) transition and 3.7 times 10-6 for the η c(2S) transition. The 90 % confidence level upper limits on the partial decay widths are also reported to be Γ(ψ(3770) to γη c(1S)) < 5.5 keV and Γ(ψ(3770) to γη c(2S)) < 29.4. With about seven times larger integrated luminosity than used previously, these results lower the upper limits by approximately a factor of three and two for the η c(1S) and η c(2S) transitions, respectively.
Sources
Related papers
- DrSai: an agentic AI system toward autonomous physics analysis workflows at BESIII
- CANTON- mu Proposal: A Next-Generation Muon g-2 Measurement at Sub-0.1 ppm Precision
- Search for dark photon dark matter near 1,meV using a superconducting mixer
- ANUBIS detector and its sensitivity to neutral long-lived particles
- How to scale your HEP ML models: A recipe for robust architecture comparisons at scale
- Search for the 16 O(ppp) to 13 C pi+ pi+ e+ Decay Mode in Super-Kamiokande Using Machine Learning Techniques