arXiv subjects

Feng Feng

Publications and source records attributed to Feng Feng.

At least 19 recordsLinked to original sources

arXiv2026

Next-to-next-to-leading-order QCD corrections to ${}^3S_1^{(8)}$ gluon fragmentation function for quarkonium

We present the first computation of the next-to-next-to-leading-order (NNLO) QCD corrections to the ${}^3S_1^{(8)}$ gluon fragmentation function for quarkonium within the nonrelativistic QCD (NRQCD) factorization framework, accurate to the lowest order in the velocity expansion. The calculation is performed with high numerical precision and encompasses both polarized and unpolarized cases. We find that the NNLO corrections are positive and substantial across most of the $z$ region. Furthermore, the logarithmic singularities near the endpoint $z\to 1$ are fully reconstructed, providing essential inputs for future threshold resummation beyond leading-logarithmic accuracy. Combined with threshold-resummed formulas in the large-$z$ region, our results yield phenomenologically viable inputs for the $^3S_1^{(8)}$ gluon fragmentation function. This enables a more reliable description of large-$p_T$ $J/\psi$ ($\psi'$) and $\chi_{cJ}$ production and polarization at hadron colliders, representing a crucial step toward a definitive test of the color-octet mechanism.

hep-ph↗

arXiv2026

Deeply virtual pion production through two-loop order

Deeply virtual meson production (DVMP) is among the most prominent channels to extract the nucleon's generalized parton distributions (GPDs) at $ep$ scattering facilities such as {\tt JLab} and the upcoming {\tt EIC/EicC} experiments, which plays a vital role in unravelling the three-dimensional internal structure of nucleon. In this work we calculate for the first time the next-to-next-to-leading order (NNLO) QCD radiative corrections to the DV$\pi$P processes $\gamma_L^* p\to \pi^+ n$ and $\gamma_L^* p\to \pi^0 p$ in the generalized Bjorken limit $Q^2\gg \vert t\vert, \Lambda_{\text{QCD}}^2$, accurate at the leading twist within collinear factorization framework. The impact of the two-loop QCD corrections appears to be positive and substantial, including which considerably improves the agreement between the perturbative QCD prediction and the available {\tt JLab} data. In addition to the differential longitudinal DV$\pi$P cross section, we also study the impact of the two-loop QCD corrections on the transverse single-spin asymmetries (TSSA) in some benchmark kinematics at {\tt JLab}, {\tt EIC} and {\tt EicC}.

hep-ph↗

arXiv2026

Q-Mask: Query-driven Causal Masks for Text Anchoring in OCR-Oriented Vision-Language Models

Optical Character Recognition (OCR) is increasingly regarded as a foundational capability for modern vision-language models (VLMs), enabling them not only to read text in images but also to support downstream reasoning in real-world visual question answering (VQA). However, practical applications further require reliable text anchors, i.e., accurately grounding queried text to its corresponding spatial region. To systematically evaluate this capability, we introduce TextAnchor-Bench (TABench), a benchmark for fine-grained text-region grounding, which reveals that both general-purpose and OCR-specific VLMs still struggle to establish accurate and stable text anchors. To address this limitation, we propose Q-Mask, a precise OCR framework built upon a causal query-driven mask decoder (CQMD). Inspired by chain-of-thought reasoning, Q-Mask performs causal visual decoding that sequentially generates query-conditioned visual masks before producing the final OCR output. This visual CoT paradigm disentangles where the text is from what the text is, enforcing grounded evidence acquisition prior to recognition and enabling explicit text anchor construction during inference. To train CQMD, we construct TextAnchor-26M, a large-scale dataset of image-text pairs annotated with fine-grained masks corresponding to specific textual elements, encouraging stable text-region correspondences and injecting strong spatial priors into VLM training. Extensive experiments demonstrate that Q-Mask substantially improves text anchoring and understanding across diverse visual scenes.

cs.CV↗

arXiv2026

Exclusive Decays of the Fully Heavy Tetraquarks into Light Mesons

In this work, we investigate the exclusive decays of the fully heavy tetraquark states $T_{4c,b}$ into light mesons, specifically $\pi$ and $K$, using the framework of Non-Relativistic QCD (NRQCD) and collinear QCD factorization for hard exclusive processes. We estimate the decay widths to be $10^{-9}$ GeV and $10^{-14}$ GeV for the decays $T_{4c} \to \pi^+\pi^-$ and $T_{4b} \to \pi^+\pi^-$ (and similarly for $K^+K^-$), respectively. The branching ratio for $T_{4c} \to \pi^+\pi^-(K^+K^-)$ is on the order of $10^{-8}$, making it currently unobservable in existing experiments. The factorization of $T_{4c,b}$ into light hadrons shares similarities with the decay $J/\psi \to p\bar{p}$. However, unlike the latter process, the decay $T_{4c,b} \to \pi^+\pi^-(K^+K^-)$ exhibits unique features that arise only in processes involving multiple incoming or outgoing particles. One such feature is the necessity of maintaining the $i\varepsilon$-prescription for the denominators or propagators due to the divergences in the kinematic region of interest. Employing the sector decomposition method, along with the aid of the Cheng-Wu theorem and the {\tt QHull} program, we present a systematic approach to handle the convolutions and phase-space integrations. This method can also be extended to similar processes, such as $T_{4c,b} \to p\bar{p}$, as well as to phase-space integrations where each denominator is a linear combination of integration variables.

hep-ph↗

arXiv2025

Mixed Electroweak-QCD Corrections to $H\to \gamma\gamma$

We present for the first time the complete three-loop mixed electroweak-QCD ($\mathcal{O}(\alpha\alpha_s)$) corrections for the decay channel $H \to \gamma\gamma$, by implementing three different on-shell $\alpha$ schemes in computing the electroweak correction. Our studies indicate that the $\mathcal{O}(\alpha_s)$ correction amounts to approximately $1.7\%$ of the leading-order prediction for the diphoton width, while the $\mathcal{O}(\alpha)$ correction varies from $-4.8\%$ to $1.4\%$ depending on the specific $\alpha$ scheme. The three-loop mixed electroweak-QCD correction may reach $0.6\%$, $0.5\%$, and $0.2\%$ of the LO diphoton width in $\alpha(0)$, $\alpha(M_Z)$, and $G_\mu$ schemes, respectively, which is much more significant than the less-than-$0.1\%$ contribution from the three-loop QCD correction. It is also worth noting that the inclusion of the ${\cal O}(\alpha\alpha_s)$ correction significantly reduces the scheme dependence of the partial width from $0.6$ keV at leading order down to $0.03$ keV. The state-of-the-art Standard Model predictions are $\Gamma[H \to \gamma\gamma] = 9.389\div 9.420$ keV, providing a valuable theoretical benchmark for future Higgs factory collider program.

hep-ph↗

arXiv2025

Optimized QCD two-loop correction to exclusive double $J/\psi$ production at B factories

We report the calculation of the process $e^+ e^- \to J/\psi J/\psi$ up to next-to-next-to-leading order (NNLO) at a center-of-mass (CM) energy of $\sqrt{s}=10.58$ GeV. We employ an improved NRQCD factorization approach, decomposing the amplitude into photon-fragmentation and non-fragmentation components. The fragmentation contribution is determined using the measured $J/\psi$ decay constant, while the interference and non-fragmentation parts are computed at NNLO in $\alpha_s$ and lowest order in velocity. In this optimized scheme, both ${\cal O}(\alpha_s)$ and ${\cal O}(\alpha^2_s)$ corrections in the interference part are positive and exhibit good convergence. The non-fragmentation part is numerically insignificant. Our results indicate that with the projected 50 ${\rm ab}^{-1}$ dataset at \texttt{Belle 2}, the prospects for observing exclusive double $J/\psi$ production are very promising.

hep-ph↗

arXiv2025

Advanced Space Mapping Technique Integrating a Shared Coarse Model for Multistate Tuning-Driven Multiphysics Optimization of Tunable Filters

This article introduces an advanced space mapping (SM) technique that applies a shared electromagnetic (EM)-based coarse model for multistate tuning-driven multiphysics optimization of tunable filters. The SM method combines the computational efficiency of EM single-physics simulations with the precision of multiphysics simulations. The shared coarse model is based on EM single-physics responses corresponding to various nontunable design parameters values. Conversely, the fine model is implemented to delineate the behavior of multiphysics responses concerning both nontunable and tunable design parameter values. The proposed overall surrogate model comprises multiple subsurrogate models, each consisting of one shared coarse model and two distinct mapping neural networks. The responses from the shared coarse model in the EM single-physics filed offer a suitable approximation for the fine responses in the multiphysics filed, whereas the mapping neural networks facilitate transition from the EM single-physics field to the multiphysics field. Each subsurrogate model maintains consistent nontunable design parameter values but possesses unique tunable design parameter values. By developing multiple subsurrogate models, optimization can be simultaneously performed for each tuning state. Nontunable design parameter values are constrained by all tuning states, whereas tunable design parameter values are confined to their respective tuning states. This optimization technique simultaneously accounts for all the tuning states to fulfill the necessary multiple tuning state requirements. Multiple EM and multiphysics training samples are generated concurrently to develop the surrogate model. Compared with existing direct multiphysics parameterized modeling techniques, our proposed method achieves superior multiphysics modeling accuracy with fewer training samples and reduced computational costs.

eess.SP↗

arXiv2024

Confronting perturbative QCD with the hardest exclusive reactions: kaon electromagnetic form factors

Among countless channels of hard exclusive reactions, the kaon electromagnetic form factors (EMFFs) are of special interest, which have been measured up to $Q^2 \sim 50\;{\rm GeV}^2$ in the timelike domain. The kaon EMFFs thereby serve an ideal platform to critically examine the validity and effectiveness of perturbative QCD (pQCD) in accounting for hard exclusive processes. In this work we confront the pQCD predictions that incorporate the next-to-next-to-leading-order (NNLO) perturbative corrections, with the available kaon EMFFs data set from experimental measurements and from lattice predictions. The inclusion of the NNLO corrections turns out to have a substantial and positive impact. If the profiles of the kaon light-cone distribution amplitudes (LCDAs) are taken from the recent lattice QCD prediction by {\tt LPC} Collaboration, the satisfactory agreement between theory and data can be reached for both charged and neutral kaons, in both spacelike and timelike large-$Q^2$ domains.

hep-ph↗

arXiv2024

Next-to-leading-order QCD corrections to nucleon Dirac form factors

The leading-order perturbative QCD (pQCD) predictions to nucleon electromagnetic form factors were first made in late 70s. In this Letter for the first time we accomplish the calculation of the next-to-leading-order (NLO) QCD corrections to nucleon's Dirac form factors at large momentum transfer, to the leading-twist accuracy in collinear factorization approach, specifically within the Krankl and Manashov renormalization scheme. The effect of NLO perturbative corrections turns out to be positive and substantial. Taking the nucleon leading-twist light-cone-distribution amplitudes (LCDAs) determined from the recent lattice simulations as input, we find that the state-of-the-art pQCD predictions significantly underestimate the available nucleon Dirac form factors in both space-like and time-like domains. This nuisance indicates that some additional soft nonfactorizable contribution might be called for to account for the measured nucleon electromagnetic form factor data up to $Q^2\approx 30\;{\rm GeV^2}$.

hep-ph↗

arXiv2024

Next-to-leading-order electroweak correction to $H\to Z^0\gamma$

Inspired by the recent observation of the Higgs boson radiative decay into $Z^0$ by {\tt ATLAS} and {\tt CMS} Collaborations, we investigate the next-to-leading-order (NLO) electroweak correction to this rare decay process in Standard Model (SM). Implementing the on-shell renormalization scheme, we find that the magnitude of the NLO electroweak correction may reach $7\%$ of the leading order (LO) prediction, much more significant than that of the NLO QCD correction, which is merely about $0.3\%$. After incorporating the ${\cal O}(\alpha)$ correction, the predicted partial width from various $\alpha$ schemes tend to converge to each other. Including both NLO electroweak and QCD corrections, the SM prediction for the branching fraction shifts from the LO value of $(1.40-1.71)\times 10^{-3}$ to $(1.55\pm 0.06)\times 10^{-3}$, considerably lower than the measured value ${\cal B}_{\rm exp}[H\to Z^0\gamma]=(3.4\pm 1.1)\times 10^{-3}$. Resolving this alarming discrepancy clearly calls for further theoretical investigations, and, more importantly, experimental efforts from {\tt HL-LHC} and the prospective Higgs factories such as {\tt CEPC} and {\tt FCC-ee}.

hep-ph↗

arXiv2024

Producing Fully-Charmed Tetraquarks via Charm Quark Fragmentation in Colliders

Within the framework of nonrelativistic QCD (NRQCD), we calculate the fragmentation function for a charm quark into an $S$-wave fully-charmed tetraquark, denoted as $T_{4c}$. The charm-to-$T_{4c}$ fragmentation function is expressed as a sum of products of the perturbatively calculable short-distance coefficients and the nonperturbative long-distance matrix elements (LDMEs). The short-distance coefficients are ascertained through the perturbative matching procedure at lowest order in $\alpha_{s}$ expansion. The LDMEs are approximated using the $T_{4c}$ four-body wave functions at the origin, which have been evaluated by various phenomenological potential models in literature. Incorporating the celebrated QCD factorization and the charm-to-$T_{4c}$ fragmentation function, we predict the $T_{4c}$ production rate at high transverse momentum $p_T$ regime in colliders. %After implementing appropriate kinematic constraints, Both the differential distribution over $p_T$ and the integrated cross sections are predicted at the \texttt{LHC}. The cross sections for $T_{4c}$ states production can reach several femtobarns to several hundreds femtobarns, suggesting a substantial potential for $T_{4c}$ event production at the \texttt{LHC}. Additionally, we estimate for the photoproduction of $T_{4c}$ in electron-proton ($ep$) collisions. It is observed that the cross sections for these processes are moderate at the \texttt{HERA} and \texttt{EIC}, and relatively small at the \texttt{EicC}. Given the luminosities of these colliders, the prospect of detecting these fully-charmed tetraquarks at $ep$ colliders is somewhat challenging.

hep-ph↗

arXiv2023

High-order Finite-Volume Central Targeted ENO Family Scheme for Compressible Flows in Unstructured Meshes

The high-order Target ENO (TENO) scheme, known for its innovative weighting strategy, has demonstrated strong potential for complex flow predictions. This study extends the TENO weighting approach to develop non-oscillatory central TENO (CTENO and CTENOZ) family schemes for unstructured meshes. The CTENO schemes employ compact directional stencils, which increase the likelihood of finding stencils within smooth regions. The design is intentionally compact to simplify the implementation of directional stencils. An effective scale separation strategy is adopted using an ENO-like stencil selection method, which employs large central stencils in smooth regions to achieve high-order accuracy, and smaller directional stencils near discontinuities to improve shock-capturing capabilities. Extensive tests involving CWENO, TENO, CTENO, and CTENOZ schemes were conducted to assess their performance in terms of accuracy, robustness, parallel scalability, and computational efficiency. The findings indicate that the proposed CTENO and CTENOZ schemes deliver high-order precision, lower numerical dissipation, and excellent shock-capturing performance.

physics.flu-dyn↗

arXiv2023

Next-to-next-to-leading-order QCD corrections to pion electromagnetic form factors

We investigate the next-to-next-to-leading order (NNLO) QCD radiative corrections to the pion electromagnetic form factor with large momentum transfer. We explicitly verify the validity of the collinear factorization to two-loop order for this observable, and obtain the respective IR-finite two-loop hard-scattering kernel in the closed form. The NNLO QCD correction turns to be positive and significant. Incorporating this new ingredient of correction, we then make a comprehensive comparison between the finest theoretical predictions and numerous pion form factor measurements in both space-like and time-like regions. Our phenomenological analysis provides strong constraint on the second Gegenbauer moment of the pion light-cone distribution amplitude (LCDA) obtained from recent lattice QCD studies.

hep-ph↗

arXiv2023

Photoproduction of fully charmed tetraquark at electron-ion colliders

In this work, we investigate the inclusive photoproduction of the $C$-odd, $S$-wave fully charmed tetraquark at electron-ion colliders within the nonrelativistic QCD (NRQCD) factorization framework, at the lowest order in velocity and $\alpha_s$. The value of the NRQCD long-distance matrix element is estimated from two phenomenological potential models. Our studies reveal that the photoproduction of the $1^{+-}$ fully charmed tetraquark may be difficult to observe at HERA and the EicC; nevertheless, its observation prospect at the EIC appears to be bright.

hep-ph↗

arXiv2023

Two-Loop QCD Corrections to C even Bottomonium Exclusive Decays to Double $J/\psi$

In the framework of nonrelativistic QCD (NRQCD) factorization, we compute both the polarized and the unpolarized decay widths for the processes $\eta_b(\chi_{bJ})\to J/\psi J/\psi$, accurate up to next-to-next-to-leading-order (NNLO) in $\alpha_s$. For the first time, we confirm that the NRQCD factorization does hold at NNLO for the process involving triple quarkonia. We find the radiative corrections are considerable. In particular for $\chi_{b2}$, both $\mathcal{O}(\alpha_s)$ and $\mathcal{O}(\alpha_s^2)$ corrections are sizable and negative, and can significantly reduce the leading order prediction. At NNLO, the branching fractions are $8.2\times 10^{-7}$, $6.2\times 10^{-6}$, $7.2\times 10^{-7}$ and $2.7\times 10^{-6}$ for $\eta_b$, $\chi_{b0}$, $\chi_{b1}$ and $\chi_{b2}$ decay, respectively. Our theoretical predictions are consistent with the upper limits measured by the {\tt Belle} Collaboration. Moreover, we investigate the dependence of the theoretical predictions on the ratio of the charm quark mass and the bottom quark mass. By fixing $m_b$ and varying $m_c$ from $1.25$ to $1.9$ GeV, we find the branching fraction can change a factor of $2$, $3$, and $6$ for $\eta_b$, $\chi_{b0}$, and $\chi_{b1}$, respectively. In the phenomenological analysis, with the integrate luminosity $\mathcal{L}=100\,{\rm fb}^{-1}$, we expect about $(5-10)\times 10^3$ $\eta_b(\chi_{bJ})\to J/\psi J/\psi\to \ell \bar{\ell}\ell \bar{\ell}$ events produced at the {\tt LHC}, thus it might be hopeful to search for these processes. On the other hand, there are less than $100$ $\eta_b(\chi_{bJ})\to J/\psi J/\psi$ signal events at the B factory, so it seems the experimental measurements on these channels are quite challenging based on current dataset.

hep-ph↗

arXiv2023

Electromagnetic and hadronic decay of fully heavy tetraquark

In this study, we compute the electromagnetic and hadronic decay widths of the S-wave fully heavy tetraquark $T_{4Q}$ ($Q=c$ or $b$) at lowest order in $\alpha_s$ and $v$, in the framework of nonrelativistic QCD. The short-distance coefficients are determined through the standard procedure of matching. The nonperturbative long-distance matrix elements are related to the phenomenological four-body Schr\"odinger wave functions at the origin, whose values are taken from literature. The branching fractions are predicted to be around $10^{-4}$ and $10^{-7}-10^{-6}$ for the $T_{4c}$ hadronic decay and electromagnetic decay, respectively. Combing our results with the $T_{4c}$ production cross sections at the LHC, we also predict the event numbers for various decay channels. With integrated luminosity $\mathcal{L}=100 \,{\rm fb}^{-1}$, it is expected that the event numbers can reach $10^3$ for $T_{4c}\to \gamma\gamma$, and $10^6$ for $T_{4c}\to {\rm LH}$, at the LHC. The detecting prospect is promising. In addition, the decay widths of $T_{4b}$ are estimated based on simple dimensional analysis as well as velocity scaling rule.

hep-ph↗

arXiv2023

Optimized ${\cal O}(\alpha_s^2)$ correction to exclusive double $J/\psi$ production at $B$ factories

The failure of observing the $e^+ e^- \to J/\psi+J/\psi$ events at $B$ factories to date is often attributed to the significant negative order-$\alpha_s$ correction. In this work we compute the ${\cal O}(\alpha^2_s)$ correction to this process for the first time. The magnitude of the next-to-next-to-leading order (NNLO) perturbative correction is substantially negative so that the standard NRQCD prediction would suffer from an unphysical, negative cross section. This dilemma may be traced in the fact that the bulk contribution of the fixed-order radiative corrections stems from the perturbative corrections to the $J/\psi$ decay constant. We thus implement an improved NRQCD factorization framework, by decomposing the amplitude into the photon-fragmentation piece and the non-fragmentation piece. With the measured $J/\psi$ decay constant as input, which amounts to resumming a specific class of radiative and relativistic corrections to all orders, the fragmentation-induced production rate can be predicted accurately and serves a benchmark prediction. The non-fragmentation type of the amplitude is then computed through NNLO in $\alpha_s$ and at lowest order in velocity. Both the ${\cal O}(\alpha_s)$ and ${\cal O}(\alpha^2_s)$ corrections in the interference term become positive and exhibit a decent convergence behavior. Our finest prediction is $\sigma(e^+ e^- \to J/\psi+J/\psi)= 2.13^{+0.30}_{-0.06}$ fb at $\sqrt{s}=10.58$ GeV. With the projected integrated luminosity of 50 ${\rm ab}^{-1}$, the prospect to observe this exclusive process at \texttt{Belle} 2 experiment appears to be bright.

hep-ph↗

arXiv2023

Inclusive production of fully-charmed tetraquarks at LHC

The $X(6900)$ resonance, originally discovered by the \texttt{LHCb} collaboration and later confirmed by both \texttt{ATLAS} and \texttt{CMS} experiments, has sparked broad interests in the fully-charmed tetraquark states. Relative to the mass spectra and decay properties of fully-heavy tetraquarks, our knowledge on their production mechanism is still rather limited. In this work we investigate the inclusive production of fully-charmed $S$-wave tetraquarks at \texttt{LHC} within the nonrelativistic QCD (NRQCD) factorization framework. The partonic cross sections are computed at lowest order in $\alpha_s$ and velocity, while the long-distance NRQCD matrix elements are estimated from phenomenological potential models. We predict the differential $p_T$ spectra of various fully-charmed $S$-wave tetraquarks at the \texttt{LHC}, and compare with the results predicted from the fragmentation mechanism at large $p_T$ end.

hep-ph↗