arXiv subjects

Yongchao Zhang

Publications and source records attributed to Yongchao Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Probing Neutrinophilic Scalars in Muon Decays

The non-standard self-interactions of neutrinos could be induced by (light) scalars that couple primarily to neutrinos. In this work, we investigate the effects of neutrinophilic scalars $ϕ$ on muon decays at the tree, 1-loop and 2-loop levels, considering both Dirac and Majorana neutrinos. The most phenomenologically interesting processes are the 1-loop corrections of $ϕ$ to the SM muon decay and the four-body muon decays $μ^- \to e^- ν_μν_μϕ,\, e^- \barν_e \barν_e ϕ$, both induced by the coupling $h_{eμ}$ of $ϕ$ to Majorana neutrinos. The infrared divergences in these processes cancel with each other, yielding an infrared-safe inclusive decay width. The precise $G_F$ measurements constrain the coupling $|h_{eμ}|$ down to roughly $0.027$, with the scalar mass $m_ϕ$ constrained up to roughly $5.9$ TeV. For $m_ϕ\gtrsim {\cal O}({\rm GeV})$, the $G_F$ measurements are more stringent than other existing laboratory, astrophysical and cosmological constraints. Some other $ϕ$-induced processes are also possible, which are, however, heavily suppressed by the small couplings, tiny neutrino masses and/or the loop factor.

hep-ph↗

arXiv2026

Probing 5.49 MeV Solar Axions at Xenon Experiments

The monochromatic 5.49 MeV solar axions induced by the isovector coupling $g_{3aN}$ can be searched for at the dark matter direct detection experiments. In this paper we estimate the prospects of the relevant axion couplings for axion mass $m_{a} <$ 1 MeV with xenon targets. Given the axion-electron coupling $g_{ae}$, the signal is dominated by the axion-induced $e^+ e^-$ pair production whose cross section is largely enhanced when the axion mass approaches twice of the electron mass. Furthermore, the cross section depends on the atomic number squared $Z^2$. This allows the next-generation xenon experiments to surpass the current Borexino constraints and provide sensitivities competitive with those of the large neutrino detectors such as JUNO and Hyper-Kamiokande. With an exposure of 200 and 1000 ton$\cdot$yr, the couplings $|g_{3aN} g_{ae}|$ can be probed down to $1.59\times10^{-14}$ and $7.12\times10^{-15}$ at 90% C.L., respectively. If the axion couples to photons, the projected sensitivities on $|g_{3aN}g_{aγ}|$ can touch down to $6.76\times10^{-12}$ GeV$^{-1}$ and $3.02\times10^{-12}$ GeV$^{-1}$, respectively.

hep-ph↗

arXiv2026

Probing Light Dark Particles in Neutrino Scattering Experiments

In this work we investigate the production of a dark fermionic particle $χ$ in the neutrino scattering experiments. In the framework of effective field theory, such process can be induced by the effective four-fermion interactions involving neutrinos, the dark particle $χ$ and standard model particles. We perform a comprehensive analysis of all possible Lorentz structures, considering representative neutrino experiments with distinct neutrino sources and target particles. In particular, we examine the constraints on the effective couplings for the neutrino-nucleus scattering by the latest COHERENT CsI and CONUS+ data, as well as the prospects at the DUNE near detector from neutrino-electron scattering. It turns out the current COHERENT and CONUS+ constraints on the cutoff scales are less stringent than those from the existing Large Hadron Collider data and the SN1987A observations. However, the DUNE near detector could probe the cutoff scales beyond the existing CHARM II and LEP limits up to roughly 1 TeV, for the dark particle mass up to roughly 50 MeV. Our results demonstrate the complementarity between neutrino experiments and collider searches in probing the dark sector physics.

hep-ph↗

arXiv2026

New Constraints on Neutrino-Dark Matter Interactions: A Comprehensive Analysis

We present a comprehensive analysis of the interactions of neutrinos with the dark sector within the simplified model framework. We first derive the exact analytic formulas for the differential scattering cross sections of neutrinos with scalar, fermion, and vector dark matter (DM) for light dark sector models with mediators of different types. We then implement the full catalog of constraints on the parameter space of the neutrino-DM and neutrino-mediator couplings and masses, including cosmological and astrophysical bounds coming from Big Bang Nucleosynthesis, Cosmic Microwave Background, DM and neutrino self-interactions, DM collisional damping, and astrophysical neutrino sources, as well as laboratory constraints from 3-body meson decays and invisible $Z$ decays. We find that most of the benchmarks in the DM mass-coupling plane adopted in previous studies to get an observable neutrino-DM interaction effect are actually ruled out by a combination of the above-mentioned constraints, especially the laboratory ones which are robust against astrophysical uncertainties and independent of the cosmological history. To illustrate the consequences of our new results, we take the galactic supernova neutrinos in the MeV energy range as a concrete example and highlight the difficulties in finding any observable effect of neutrino-DM interactions. Finally, we identify new benchmark points potentially promising for future observational prospects of the attenuation of the galactic supernova neutrino flux and comment on their implications for the detection prospects in future large-volume neutrino experiments such as JUNO, Hyper-K, and DUNE. We also comment on the ultraviolet-embedding of the effective neutrino-DM couplings.

hep-ph↗

arXiv2026

Loop-Level Lepton Flavor Violation and Diphoton Signals in the Minimal Left-Right Symmetric Model

The left-right symmetric model (LRSM) could not only restore parity of the weak interaction, but also provide natural explanations of the tiny active neutrino masses via the seesaw mechanisms. The $SU(2)_R$-breaking scalar $H_3$ can induce lepton flavor violating (LFV) effects in the minimal version of LRSM at the 1-loop order, originating from the mixing of heavy right-handed neutrinos (RHNs). If $H_3$ is light, say below the GeV scale, it will lead to rich signals, e.g. the LFV muon and tauon decays $\ell_β\to \ell_α+ X$ ($X$ being either visible or invisible final states) and the anomalous supernova signatures. Combined with the diphoton coupling of $H_3$, and recasting the existing constraints onto the light $H_3$ scenario, the right-handed scale $v_R$ is excluded up to $2\times10^9$ GeV. In the future, the $v_R$ scale can be probed up to $5\times10^9$ GeV in high-precision muon experiments, if the Yukwa couplings for RHN masses are of order one and the RHN mixing is maximal, and further up to $6\times10^{11}$ GeV by supernova observations, reaching the non-resonant leptogenesis scale in the LRSM.

hep-ph↗

arXiv2026

Looking for Lights from the Darkness: Signals from MeV-scale Solar Axion-like Particles

The axion-like particles $a$ can be produced in the Sun via the process of $p + D \to {}^3{\rm He} +a$, with mass up to 5.5 MeV. The photons in the subsequent decay $a \to γγ$ can deviate significantly from the Sun, or even from roughly the opposite direction of the Sun. The nontrivial angular and spectral distributions of such photons enable us new methods to detect the {\it lights from the darkness}. In this letter, we consider both the space detection and terrestrial experiments at the South Pole. As a result of the two-body decay and the geometric effects, there exists a critical height for the terrestrial experiments, below which there is no photon for some regions of the parameter space. With the sensitivities of $10^{-16}$ ($10^{-17}$) erg cm$^{-2}$ s$^{-1}$ for the MeV-scale photons in future space and terrestrial experiments, the coupling $g_{aγ}$ of $a$ to photons can be probed up to $3\times10^{-12}$ ($1\times10^{-12}$) GeV$^{-1}$, well surpassing the current supernova limits.

hep-ph↗

arXiv2026

Flash from the Past: New Gamma-Ray Constraints on Light CP-even Scalar from SN1987A

We derive new constraints on light CP-even scalars using old gamma-ray observations in the direction of SN1987A by the Solar Maximum Mission (SMM) satellite. Light scalars can be abundantly produced in the supernova core via the nucleon bremsstrahlung process, can stream out of the supernova-environment and decay into photons -- either primary photons or secondary photons from lepton-antilepton pairs -- thus leading to a gamma-ray signal. From the non-observation of excess photon flux by SMM after the detection of the neutrino burst from SN1987A, we set new constraints on the mixing angle of the CP-even scalar with the Standard Model Higgs boson.

hep-ph↗

arXiv2025

Conceptual Design of the Muonium-to-Antimuonium Conversion Experiment (MACE)

The spontaneous conversion of muonium to antimuonium is one of the interesting charged lepton flavor violation phenomena offering a sensitive probe of potential new physics and serving as a tool to constrain the parameter space beyond the Standard Model. The Muonium-to-Antimuonium Conversion Experiment (MACE) is designed to utilize a high-intensity muon beam, a Michel electron magnetic spectrometer, a positron transport system, and a positron detection system, to either discover or constrain this rare process with a conversion probability of $\mathcal{O}(10^{-13})$. This article presents an overview of the theoretical framework as well as a detailed description of the experimental design for the search for muonium-to-antimuonium conversion.

hep-ex↗

arXiv2025

BrainHGT: A Hierarchical Graph Transformer for Interpretable Brain Network Analysis

Graph Transformer shows remarkable potential in brain network analysis due to its ability to model graph structures and complex node relationships. Most existing methods typically model the brain as a flat network, ignoring its modular structure, and their attention mechanisms treat all brain region connections equally, ignoring distance-related node connection patterns. However, brain information processing is a hierarchical process that involves local and long-range interactions between brain regions, interactions between regions and sub-functional modules, and interactions among functional modules themselves. This hierarchical interaction mechanism enables the brain to efficiently integrate local computations and global information flow, supporting the execution of complex cognitive functions. To address this issue, we propose BrainHGT, a hierarchical Graph Transformer that simulates the brain's natural information processing from local regions to global communities. Specifically, we design a novel long-short range attention encoder that utilizes parallel pathways to handle dense local interactions and sparse long-range connections, thereby effectively alleviating the over-globalizing issue. To further capture the brain's modular architecture, we designe a prior-guided clustering module that utilizes a cross-attention mechanism to group brain regions into functional communities and leverage neuroanatomical prior to guide the clustering process, thereby improving the biological plausibility and interpretability. Experimental results indicate that our proposed method significantly improves performance of disease identification, and can reliably capture the sub-functional modules of the brain, demonstrating its interpretability.

cs.LG↗

arXiv2025

New Laboratory Constraints on Neutrinophilic Mediators

Neutrinophilic mediators are well-motivated messenger particles that can probe some of the least known sectors of fundamental physics involving nonstandard interactions of neutrinos with themselves and potentially with dark matter. In particular, light mediators coupling to the active neutrinos will induce new decay modes of the Standard Model mesons (e.g., $π^\pm, K^\pm \to \ell^\pm + ν+ ϕ$), charged leptons (e.g., $τ^\pm \to π^\pm + ν+ ϕ$), and gauge bosons (e.g., $Z \to ν+ \barν+ ϕ$). A common lore is that these decays suffer from infrared divergences in the limit of the vanishing mediator mass, i.e., $m_ϕ\to 0$. Here, we show that including the 1-loop contributions of these mediators to the standard 2-body decays (e.g., $π^\pm,\,K^\pm \to \ell^\pm + ν$, etc.), the infrared divergence from the 3-body decay cancels out exactly by virtue of the Kinoshita-Lee-Nauenberg theorem. Including these cancellation effects, we then update the existing laboratory constraints on neutrinophilic scalar mediators, thereby extending the limits far beyond the decaying parent particle mass and excluding a wider range of parameter space. These new ``physical'' limits derived here have significant implications for the future detection prospects of nonstandard neutrino (self-)interactions.

hep-ph↗

arXiv2025

Heavy neutrino mixing prospects at hadron colliders: a machine learning study

We apply machine learning to the searches of heavy neutrino mixing in the inverse seesaw in the framework of left-right symmetric model at the high-energy hadron colliders. The Majorana nature of heavy neutrinos can induce the processes $pp \to W_R^\pm \to \ell_α^\pm N \to \ell_α^\pm \ell_β^{\mp,\,\pm} jj$, with opposite-sign (OS) and same-sign (SS) dilepton and two jets in the final state. The distributions of the charged leptons $\ell = e ,\, μ$ and jets and their correlations are utilized as input for machine learning analysis. It is found that for both the OS and SS processes, XGBoost can efficiently distinguish signals from the standard model backgrounds. We estimate the sensitivities of heavy neutrino mass $m_N$ and their mixing in the OS and SS $ee$, $μμ$ and $eμ$ final states at $\sqrt{s} = 14$ TeV, 27 TeV and 100 TeV. It turns out that the heavy neutrinos can be probed up to 17.1 TeV and 19.5 TeV in the OS and SS channels, respectively. The sine of the mixing angle of heavy neutrinos can be probed up to the maximal value of $\sqrt2/2$ and 0.69 in the OS and SS channels, respectively.

hep-ph↗

arXiv2024

Error Analysis and Numerical Algorithm for PDE Approximation with Hidden-Layer Concatenated Physics Informed Neural Networks

We present the hidden-layer concatenated physics informed neural network (HLConcPINN) method, which combines hidden-layer concatenated feed-forward neural networks, a modified block time marching strategy, and a physics informed approach for approximating partial differential equations (PDEs). We analyze the convergence properties and establish the error bounds of this method for two types of PDEs: parabolic (exemplified by the heat and Burgers' equations) and hyperbolic (exemplified by the wave and nonlinear Klein-Gordon equations). We show that its approximation error of the solution can be effectively controlled by the training loss for dynamic simulations with long time horizons. The HLConcPINN method in principle allows an arbitrary number of hidden layers not smaller than two and any of the commonly-used smooth activation functions for the hidden layers beyond the first two, with theoretical guarantees. This generalizes several recent neural-network techniques, which have theoretical guarantees but are confined to two hidden layers in the network architecture and the $\tanh$ activation function. Our theoretical analyses subsequently inform the formulation of appropriate training loss functions for these PDEs, leading to physics informed neural network (PINN) type computational algorithms that differ from the standard PINN formulation. Ample numerical experiments are presented based on the proposed algorithm to validate the effectiveness of this method and confirm aspects of the theoretical analyses.

math.NA↗

arXiv2024

First Constraints on the Photon Coupling of Axion-like Particles from Multimessenger Studies of the Neutron Star Merger GW170817

We use multimessenger observations of the neutron star merger event GW170817 to derive new constraints on axion-like particles (ALPs) coupling to photons. ALPs are produced via Primakoff and photon coalescence processes in the merger, escape the remnant and decay back into two photons, giving rise to a photon signal approximately along the line-of-sight to the merger. We analyze the spectral and temporal information of the ALP-induced photon signal, and use the Fermi-LAT observations of GW170817 to derive our new ALP constraints. We also show the improved prospects with future MeV gamma-ray missions, taking the spectral and temporal coverage of Fermi-LAT as an example.

hep-ph↗

arXiv2024

Explanation of the 95 GeV $γγ$ and $b\bar{b}$ excesses in the Minimal Left-Right Symmetric Model

We propose a simple interpretation of the $γγ$ excesses reported by both CMS and ATLAS groups at 95 GeV together with the LEP excess in the $Zb\bar{b}$ channel around the same mass in terms of a neutral scalar field in the minimal left-right symmetric model (LRSM). We point out that the scalar field which implements the seesaw mechanism for neutrino masses has all the right properties to explain these observations, without introducing any extra scalar fields. The key point is that this scalar particle is hardly constrained because it couples only to heavy right-handed particles. As a result, the diphoton decay mode receives contributions from both mixing with the Standard Model (SM) Higgs and the heavy charged bosons in the LRSM, depending on the $SU(2)_R\times U(1)_{B-L}$ symmetry breaking scale $v_R$. The complete allowed parameter space for explaining the 95 GeV excesses in this model can be probed with the high-precision measurements of the SM Higgs mixing with other scalars at the high-luminosity LHC and future Higgs factories.

hep-ph↗

arXiv2023

Light scalar explanation for 18 TeV GRB 221009A

Recent astrophysical transient Swift J1913.1+1946 may be associated with the $γ$-ray burst GRB 221009A. The redshift of this event is $z\simeq 0.151$. Very high-energy $γ$-rays (up to 18 TeV) followed the transient and were observed by LHAASO, additionally Carpet-2 detected a photon-like air shower of 251 TeV. Photons of such high energy are expected to readily annihilate with the diffuse extragalactic background light (EBL) before reaching Earth. If the $γ$-ray identification and redshift measurements are correct, new physics could be necessary to explain these measurements. This letter provides the first CP-even scalar explanation of the most energetic 18 TeV event reported by LHAASO. In this minimal scenario, the light scalar singlet $S$ mixes with the Standard Model (SM) Higgs boson $h$. The highly boosted $S$ particles are produced in the GRB and then undergo the radiative decay di-photon $S\rightarrow γγ$ while propagating to Earth. The resulting photons may thus be produced at a remote region without being nullified by the EBL. Hence, the usual exponential reduction of $γ$-rays is lifted due to an attenuation that is inverse in the optical depth, which becomes much larger due to the scalar carriers.

hep-ph↗

arXiv2023

Error Analysis of Physics-Informed Neural Networks for Approximating Dynamic PDEs of Second Order in Time

We consider the approximation of a class of dynamic partial differential equations (PDE) of second order in time by the physics-informed neural network (PINN) approach, and provide an error analysis of PINN for the wave equation, the Sine-Gordon equation and the linear elastodynamic equation. Our analyses show that, with feed-forward neural networks having two hidden layers and the $\tanh$ activation function, the PINN approximation errors for the solution field, its time derivative and its gradient field can be effectively bounded by the training loss and the number of training data points (quadrature points). Our analyses further suggest new forms for the training loss function, which contain certain residuals that are crucial to the error estimate but would be absent from the canonical PINN loss formulation. Adopting these new forms for the loss function leads to a variant PINN algorithm. We present ample numerical experiments with the new PINN algorithm for the wave equation, the Sine-Gordon equation and the linear elastodynamic equation, which show that the method can capture the solution well.

math.NA↗

arXiv2022

Constraints on Neutron-Mirror-Neutron Oscillation from Neutron Star Cooling

We address a method of limiting neutron-mirror neutron mixing ($ε_{nn'}$) by analyzing its effect on neutron star (NS) heating. This method employs observational bounds on the surface temperature of NSs to constrain $ε_{nn'}$. It has been suggested that the bound obtained this way is so stringent that it would exclude any discovery of $n-n'$ oscillation in the currently planned terrestrial experiments at various laboratories. This conclusion motivated us to critically analyze this suggestion in more detail. In this note, we point out a very interesting new effect present in nearly exact mirror models, which can significantly affect this bound. The new element is that in nearly exact mirror models there is the mirror analog of $β$ decay, i.e. $n' \to p' + e' + \barν'_e$, which creates a cloud of mirror particles $n'$, $p'$, $e'$, $D'$ and He$'$ inside the NS. The resulting $e'$ can "rob" the energy generated by the $n \to n'$ transition from the NS, via $e-e'$ scattering enabled by the presence of a (minute) millicharge in mirror particles. Such a tiny millicharge on mirror particles is highly likely in these models. This results in energy being emitted as unobserved mirror photons via fast mirror bremsstrahlung, whose effect is to relax the stringent bounds on $ε_{nn'}$.

hep-ph↗

arXiv2022

Improving heavy Dirac neutrino prospects at future hadron colliders using machine learning

In this work, by using the machine learning methods, we study the sensitivities of heavy pseudo-Dirac neutrino $N$ in the inverse seesaw at the high-energy hadron colliders. The production process for the signal is $pp \to \ell N \to 3 \ell + E_T^{\rm miss}$, while the dominant background is $p p \to W Z \to 3 \ell + E_T^{\rm miss}$. We use either the Multi-Layer Perceptron or the Boosted Decision Tree with Gradient Boosting to analyse the kinematic observables and optimize the discrimination of background and signal events. It is found that the reconstructed $Z$ boson mass and heavy neutrino mass from the charged leptons and missing transverse energy play crucial roles in separating the signal from backgrounds. The prospects of heavy-light neutrino mixing $|V_{\ell N}|^2$ (with $\ell = e,\,μ$) are estimated by using machine learning at the hadron colliders with $\sqrt{s}=14$ TeV, 27 TeV, and 100 TeV, and it is found that $|V_{\ell N}|^2$ can be improved up to ${\cal O} (10^{-6})$ for heavy neutrino mass $m_N = 100$ GeV and ${\cal O} (10^{-4})$ for $m_N = 1$ TeV.

hep-ph↗