Search

We introduce LLaVA-MoD, a novel framework designed to enable the efficient training of small-scale Multimodal Language Models (s-MLLM) by distilling knowledge from large-scale MLLM (l-MLLM). Our approach tackles two fundamental challenges in MLLM distillation. First, we optimize the network structure of s-MLLM by integrating a sparse Mixture of Experts (MoE) architecture into the language model, striking a balance between computational efficiency and model expressiveness. Second, we propose a progressive knowledge transfer strategy to ensure comprehensive knowledge migration. This strategy begins with mimic distillation, where we minimize the Kullback-Leibler (KL) divergence between output distributions to enable the student model to emulate the teacher network's understanding. Following this, we introduce preference distillation via Direct Preference Optimization (DPO), where the key lies in treating l-MLLM as the reference model. During this phase, the s-MLLM's ability to discriminate between superior and inferior examples is significantly enhanced beyond l-MLLM, leading to a better student that surpasses its teacher, particularly in hallucination benchmarks. Extensive experiments demonstrate that LLaVA-MoD outperforms existing models across various multimodal benchmarks while maintaining a minimal number of activated parameters and low computational costs. Remarkably, LLaVA-MoD, with only 2B activated parameters, surpasses Qwen-VL-Chat-7B by an average of 8.8% across benchmarks, using merely 0.3% of the training data and 23% trainable parameters. These results underscore LLaVA-MoD's ability to effectively distill comprehensive knowledge from its teacher model, paving the way for the development of more efficient MLLMs. The code will be available on: https://github.com/shufangxun/LLaVA-MoD.

Invasive brain-computer interfaces have garnered significant attention due to their high performance. The current intracranial stereoElectroEncephaloGraphy (sEEG) foundation models typically build univariate representations based on a single channel. Some of them further use Transformer to model the relationship among channels. However, due to the locality and specificity of brain computation, their performance on more difficult tasks, e.g., speech decoding, which demands intricate processing in specific brain regions, is yet to be fully investigated. We hypothesize that building multi-variate representations within certain brain regions can better capture the specific neural processing. To explore this hypothesis, we collect a well-annotated Chinese word-reading sEEG dataset, targeting language-related brain networks, over 12 subjects. Leveraging this benchmark dataset, we developed the Du-IN model that can extract contextual embeddings from specific brain regions through discrete codebook-guided mask modeling. Our model achieves SOTA performance on the downstream 61-word classification task, surpassing all baseline models. Model comparison and ablation analysis reveal that our design choices, including (i) multi-variate representation by fusing channels in vSMC and STG regions and (ii) self-supervision by discrete codebook-guided mask modeling, significantly contribute to these performances. Collectively, our approach, inspired by neuroscience findings, capitalizing on multi-variate neural representation from specific brain regions, is suitable for invasive brain modeling. It marks a promising neuro-inspired AI approach in BCI.

In this paper, we carry out an investigation into the semileptonic decays $D_s^+ \to K^0\ell^+\nu_\ell$ with $\ell=(e,\mu)$ by employing the QCD light-cone sum rules approach. The vector transition form factor (TFF) $f_+^{D_s^+ K^0}(q^2)$ for $D_s^+\to K^0$ decay is calculated while considering its next-to-leading order contribution. Subsequently, we briefly introduce the twist-2, 3 kaon distribution amplitudes, which are calculated by using QCD sum rules within the framework of the background field theory. At the large recoil point, the TFF has $f_+^{D_s^+ K^0}(0)=0.692_{-0.026}^{+0.027}$. Then, we extrapolate $f_+^{D_s^+ K^0}(q^2)$ to the whole physical $q^2$-region via the simplified $z(q^2,t)$-series expansion, and the behavior of TFF $f_+^{D_s^+ K^0}(q^2)$ is exhibited in the numerical results part, including the theoretical and experimental predictions for comparison. In addition, we compute the differential branching fraction $\mathcal{B}(D_s^+ \to K^0\ell^+\nu_\ell)$ with the electron and muon channels, which are expected to be $\mathcal{B}(D_s^+ \to K^0e^+\nu_e)=3.379_{-0.275}^{+0.301}\times 10^{-3}$ and $\mathcal{B}(D_s^+ \to K^0\mu^+\nu_\mu)=3.351_{-0.273}^{+0.299}\times 10^{-3}$ as well as contained other results for comparison. Our results show good agreement with the BESIII measurements and theoretical predictions. Furthermore, we present our prediction with respect to the CKM matrix element $|V_{cd}|$ by using the $\mathcal{B}(D_s^+ \to K^0e^+\nu_e)$ result from BESIII Collaboration, yielding $|V_{cd}|=0.221_{-0.010}^{+0.008}$. Finally, we provide the ratio between $D_s^+ \to K^0e^+\nu_e$ and $D_s^+ \to \eta e^+\nu_e$ channels, i.e. $\mathcal{R}_{K^0/\eta}^e=0.144_{-0.020}^{+0.028}$., Comment: 5 pages, 3 figures, comments welcome

In this paper, we aim to adapt a model at test-time using a few unlabeled data to address distribution shifts. To tackle the challenges of extracting domain knowledge from a limited amount of data, it is crucial to utilize correlated information from pre-trained backbones and source domains. Previous studies fail to utilize recent foundation models with strong out-of-distribution generalization. Additionally, domain-centric designs are not flavored in their works. Furthermore, they employ the process of modelling source domains and the process of learning to adapt independently into disjoint training stages. In this work, we propose an approach on top of the pre-computed features of the foundation model. Specifically, we build a knowledge bank to learn the transferable knowledge from source domains. Conditioned on few-shot target data, we introduce a domain prompt generator to condense the knowledge bank into a domain-specific prompt. The domain prompt then directs the visual features towards a particular domain via a guidance module. Moreover, we propose a domain-aware contrastive loss and employ meta-learning to facilitate domain knowledge extraction. Extensive experiments are conducted to validate the domain knowledge extraction. The proposed method outperforms previous work on 5 large-scale benchmarks including WILDS and DomainNet., Comment: ICLR2024, code: https://github.com/Guliisgreat/VDPG

3-dB couplers, which are commonly used in photonic integrated circuits for on-chip information processing, precision measurement, and quantum computing, face challenges in achieving robust performance due to their limited 3-dB bandwidths and sensitivity to fabrication errors. To address this, we introduce topological physics to nanophotonics, developing a framework for topological 3-dB couplers. These couplers exhibit broad working wavelength range and robustness against fabrication dimensional errors. By leveraging valley-Hall topology and mirror symmetry, the photonic-crystal-slab couplers achieve ideal 3-dB splitting characterized by a wavelength-insensitive scattering matrix. Tolerance analysis confirms the superiority on broad bandwidth of 48 nm and robust splitting against dimensional errors of 20 nm. We further propose a topological interferometer for on-chip distance measurement, which also exhibits robustness against dimensional errors. This extension of topological principles to the fields of interferometers, may open up new possibilities for constructing robust wavelength division multiplexing, temperature-drift-insensitive sensing, and optical coherence tomography applications., Comment: 20 pages, 4 figures

In the present paper, we study the properties of $\phi$-meson longitudinal leading-twist light-cone distribution amplitude $\phi_{2;{\phi}}^{\|}(x,\mu)$ by starting from a light-cone harmonic oscillator model for its wavefunction. To fix the input parameters, we derive the first ten $\xi$-moments of $\phi_{2;{\phi}}^{\|}(x,\mu)$ by using the QCD sum rules approach under the background field theory. The shape of $\phi_{2;{\phi}}^{\|}(x,\mu=2~{\rm GeV})$ tends to be a single-peak behavior, which is consistent with the latest Lattice QCD result. As an application, we derive the $D^+_s \to \phi$ transition form factors (TFFs) by using the light-cone sum rules approach. At the large recoil point, we obtain $A_1(0) = 0.512_{-0.020}^{+0.030}$, $A_2(0) = 0.402_{-0.067}^{+0.078}$, $A_0(0) = 0.596_{-0.020}^{+0.025}$ and $V(0) = 0.882_{-0.036}^{+0.040}$. As for the two typical ratios $\gamma_V$ and $\gamma_2$, we obtain $\gamma_V = 1.723_{-0.021}^{+0.023}$ and $\gamma_2 = 0.785_{-0.104}^{+0.100}$. After extrapolating those TFFs to the physically allowable region, we then obtain the transverse, longitudinal and total decay widths for semi-leptonic decay $D^+_s\to\phi\ell^+\nu_{\ell}$. Then the branching fractions are ${\cal B}(D^+_s\to \phi e^+\nu_e) = (2.367_{-0.132}^{+0.256})\times 10^{-3}$ and ${\cal B}(D^+_s\to \phi \mu^+\nu_{\mu}) = (2.349_{-0.132}^{+0.255})\times 10^{-3}$, which show good agreement with the data issued by the BESIII, the CLEO, and the BABAR Collaborations. We finally calculate $D^+_s\to\phi\ell^+ \nu_\ell$ polarization and asymmetry parameters., Comment: 32 pages, 7 figures, comments welcome

In an era of digital ubiquity, efficient resource management and decision-making are paramount across numerous industries. To this end, we present a comprehensive study on the integration of machine learning (ML) techniques into Huawei Cloud's OptVerse AI Solver, which aims to mitigate the scarcity of real-world mathematical programming instances, and to surpass the capabilities of traditional optimization techniques. We showcase our methods for generating complex SAT and MILP instances utilizing generative models that mirror multifaceted structures of real-world problem. Furthermore, we introduce a training framework leveraging augmentation policies to maintain solvers' utility in dynamic environments. Besides the data generation and augmentation, our proposed approaches also include novel ML-driven policies for personalized solver strategies, with an emphasis on applications like graph convolutional networks for initial basis selection and reinforcement learning for advanced presolving and cut selection. Additionally, we detail the incorporation of state-of-the-art parameter tuning algorithms which markedly elevate solver performance. Compared with traditional solvers such as Cplex and SCIP, our ML-augmented OptVerse AI Solver demonstrates superior speed and precision across both established benchmarks and real-world scenarios, reinforcing the practical imperative and effectiveness of machine learning techniques in mathematical programming solvers.

In the paper, we investigate the charmed meson rare decay process $D^+ \to \pi^+\nu\bar\nu$ by using QCD sum rules approach. Firstly, the pion twist-2 and twist-3 distribution amplitude $\xi$-moments $\langle\xi_{2;\pi}^n\rangle|_\mu$ up to 10th-order and $\langle \xi_{3;\pi}^{(p,\sigma),n}\rangle|_\mu$ up to fourth-order are calculated by using QCD sum rule under background field theory. After constructing the light-cone harmonic oscillator model for pion twist-2, 3 DAs, we get their behaviors by matching the calculated $\xi$-moments. Then, the $D\to \pi$ transition form factors are calculated by using QCD light-cone sum rules approach. The vector form factor at large recoil region is $f_+^{D\to\pi}(0) = 0.627^{+0.120} _{-0.080}$. By taking the rapidly $z(q^2,t)$ converging simplified series expansion, we present the TFFs and the corresponding angular coefficients in the whole squared momentum transfer physical region. Furthermore, we display the semileptonic decay process $\bar D^0 \to \pi^+ e\bar \nu_e$ differential decay widths and branching fraction with ${\cal B}(\bar D^0\to\pi^+e\bar\nu_e) = 0.308^{+0.155}_{-0.066} \times 10^{2}$. The $\bar D^0\to\pi^+e\bar\nu_e$ differential/total predictions for forward-backward asymmetry, $q^2$-differential flat terms and lepton polarization asymmetry are also given. After considering the non-standard neutrino interactions, the predictions for the $D^+ \to \pi^+ \nu\bar\nu$ branching fraction is ${\cal B}(D^+ \to \pi^+ {\nu }{\bar\nu}) = 1.85^{+0.93}_{-0.46}\times10^{-8}$., Comment: 12 pages, 9 figures

Understanding semantic content from brain activity during sleep represents a major goal in neuroscience. While studies in rodents have shown spontaneous neural reactivation of memories during sleep, capturing the semantic content of human sleep poses a significant challenge due to the absence of well-annotated sleep datasets and the substantial differences in neural patterns between wakefulness and sleep. To address these challenges, we designed a novel cognitive neuroscience experiment and collected a comprehensive, well-annotated electroencephalography (EEG) dataset from 134 subjects during both wakefulness and sleep. Leveraging this benchmark dataset, we developed SI-SD that enhances sleep semantic decoding through the position-wise alignment of neural latent sequence between wakefulness and sleep. In the 15-way classification task, our model achieves 24.12% and 21.39% top-1 accuracy on unseen subjects for NREM 2/3 and REM sleep, respectively, surpassing all other baselines. With additional fine-tuning, decoding performance improves to 30.32% and 31.65%, respectively. Besides, inspired by previous neuroscientific findings, we systematically analyze how the "Slow Oscillation" event impacts decoding performance in NREM 2/3 sleep -- decoding performance on unseen subjects further improves to 40.02%. Together, our findings and methodologies contribute to a promising neuro-AI framework for decoding brain activity during sleep.

Motivated by our previous work [Phys. Rev. D \textbf{104}, no.1, 016021 (2021)] on pionic leading-twist distribution amplitude (DA), we revisit $\rho$-meson leading-twist longitudinal DA $\phi_{2;\rho}^\|(x,\mu)$ in this paper. A model proposed by Chang based on the Dyson-Schwinger equations (DSEs) is adopted to describe the behavior of $\phi_{2;\rho}^\|(x,\mu)$. On the other hand, the $\xi$-moments of $\phi_{2;\rho}^\|(x,\mu)$ are calculated with the QCD sum rules in the framework of the background field theory. The sum rule formula for those moments are improved. More accurate values for the first five nonzero $\xi$-moments at typical scale $\mu =1, 1.4, 2, 3~{\rm GeV}$ are given, e.g., at $\mu = 1~{\rm GeV}$, \modi{$\langle\xi^2\rangle_{2;\rho}^\| = 0.220(6) $, $\langle\xi^4\rangle_{2;\rho}^\| = 0.103(4)$, $\langle\xi^6\rangle_{2;\rho}^\| = 0.066(5)$, $\langle\xi^8\rangle_{2;\rho}^\| = 0.046(4)$ and $\langle\xi^{10}\rangle_{2;\rho}^\| = 0.035(3)$}. By fitting those values with the least squares method, the DSE model for $\phi_{2;\rho}^\|(x,\mu)$ is determined. By taking the left-handed current light-cone sum rule approach, we get the transition form factor at large recoil region, {\it i.e.} $A_1(0) = 0.498^{+0.014}_{-0.012}$, $A_2(0)=0.460^{+0.055}_{-0.047}$, $V(0) = 0.800^{+0.015}_{-0.014}$, and the ratio $r_2 = 0.923^{+0.133}_{-0.119}$, $r_V = 1.607^{+0.071}_{-0.071}$. After making the extrapolation with a rapidly converging series based on $z(t)$-expansion, we present the decay width for the semileptonic decays $D\to\rho\ell^+\nu_\ell$. Finally, the branching fractions are $\mathcal{B}(D^0\to \rho^- e^+ \nu_e) = 1.889^{+0.176}_{-0.170}\pm 0.005$, $\mathcal{B}(D^+ \to \rho^0 e^+ \nu_e) = 2.380^{+0.221}_{-0.214}\pm 0.012$, $\mathcal{B}(D^0\to \rho^- \mu^+ \nu_\mu) = 1.881^{+0.174}_{-0.168}\pm 0.005$, $\mathcal{B}(D^+ \to \rho^0 \mu^+ \nu_\mu) =2.369^{+0.219}_{-0.211}\pm 0.011$., Comment: 9 pages, 3 figures

The $\eta^{(\prime)}$-mesons in the quark-flavor basis are mixtures of two mesonic states $|\eta_{q}\rangle=|\bar u u+\bar d d\rangle/\sqrt 2$ and $|\eta_{s}\rangle=|\bar s s\rangle$. In the previous work, we have made a detailed study on the $\eta_{s}$ leading-twist distribution amplitude. As a sequential work, in the present paper, we fix the $\eta_q$ leading-twist distribution amplitude by using the light-cone harmonic oscillator model for its wave function and by using the QCD sum rules within the QCD background field to calculate its moments. The input parameters of $\eta_q$ leading-twist distribution amplitude $\phi_{2;\eta_q}$ at an initial scale $\mu_0\sim 1$ GeV are then fixed by using those moments. The sum rules for the $0_{\rm th}$-order moment can also be used to fix the magnitude of $\eta_q$ decay constant, which gives $f_{\eta_q}=0.141\pm0.005$ GeV. As an application of the present derived $\phi_{2;\eta_q}$, we calculate the transition form factors $B(D)^+ \to\eta^{(\prime)}$ by using the QCD light-cone sum rules up to twist-4 accuracy and by including the next-to-leading order QCD corrections to the twist-2 part, and then fix the related CKM matrix element and the decay width for the semi-leptonic decays $B(D)^+ \to\eta^{(\prime)}\ell^+ \nu_\ell$., Comment: 16 pages, 8 figures, Comments welcome

Multi-finger grasping relies on high quality training data, which is hard to obtain: human data is hard to transfer and synthetic data relies on simplifying assumptions that reduce grasp quality. By making grasp simulation differentiable, and contact dynamics amenable to gradient-based optimization, we accelerate the search for high-quality grasps with fewer limiting assumptions. We present Grasp'D-1M: a large-scale dataset for multi-finger robotic grasping, synthesized with Fast- Grasp'D, a novel differentiable grasping simulator. Grasp'D- 1M contains one million training examples for three robotic hands (three, four and five-fingered), each with multimodal visual inputs (RGB+depth+segmentation, available in mono and stereo). Grasp synthesis with Fast-Grasp'D is 10x faster than GraspIt! and 20x faster than the prior Grasp'D differentiable simulator. Generated grasps are more stable and contact-rich than GraspIt! grasps, regardless of the distance threshold used for contact generation. We validate the usefulness of our dataset by retraining an existing vision-based grasping pipeline on Grasp'D-1M, and showing a dramatic increase in model performance, predicting grasps with 30% more contact, a 33% higher epsilon metric, and 35% lower simulated displacement. Additional details at https://dexgrasp.github.io.

In this paper, the semileptonic decays $D_s^+ \to \pi^0\ell^+ \nu_\ell$ with $\ell=(e,\mu)$ are investigated by using the light-cone sum rule approach. Firstly, the neutral meson mixing scheme between $\pi^0$, $\eta$, $\eta^\prime$ and pseudoscalar gluonium $G$ is discussed in a unified way, which leads to the direct connection between two different channels for $D_s^+\to \pi^0\ell^+\nu_\ell$ and $D_s^+ \to \eta\ell^+\nu_\ell$ by the $\pi^0-\eta$ mixing angle. Then we calculated the $D_s\to \pi^0$ transition form factors (TFFs) within QCD light-cone sum rule approach up to next-to-leading order correction. At the large recoil point, we have $f_+^{D_s^+\pi^0}(0)=0.0113_{-0.0019}^{+0.0024}$ and $f_-^{D_s^+\pi^0}(0)=0.0020_{-0.0009}^{+0.0008}$. Furthermore, the TFFs are extrapolated to the whole physical $q^2$-region by using the simplified $z(q^2)$-series expansion. The behaviors of TFFs and related three angular coefficient functions $a_{\theta_\ell}(q^2)$, $b_{\theta_\ell}(q^2)$ and $c_{\theta_\ell}(q^2)$ are given. The differential decay widths for $D_s^+ \to \pi^0\ell^+ \nu_\ell$ with respect to $q^2$ and $\cos\theta_\ell$ are presented, and also lead to the branching fractions ${\cal B}(D_s^+\to \pi ^0e^+\nu_e) =2.60_{-0.51}^{+0.57}\times 10^{-5}$ and ${\cal B}(D_s^+\to \pi ^0\mu^+\nu _\mu )= 2.58_{-0.51}^{+0.56}\times 10^{-5}$. These results show well agreement with the recent BESIII measurements and theoretical predictions. Then the differential distributions and integrated predictions for three angular observables, {\it i.e.} forward-backward asymmetries, $q^2$-differential flat terms and lepton polarization asymmetries are given separately. Lastly, we estimate the ratio for different decay channels ${\cal R}_{\pi ^0/\eta}^{\ell}=1.108_{-0.071}^{+0.039}\times 10^{-3}$., Comment: 10 pages, 5 figures. Matches version accepted for publication in PRD

HS 2325+8205 is a long-period eclipsing dwarf nova with an orbital period above the period gap (Porb>3 h) and is reported to be a Z Cam-type dwarf nova. Based on the photometry of the Transiting Exoplanet Survey Satellite (TESS), the light variation and the quasi-periodic oscillation (QPOs) of HS 2325+8205 are studied. Using Continuous Wavelet Transform (CWT), Lomb-Scargle Periodogram (LSP), and sine fitting methods, we find for the first time that there is a QPOs of ~ 2160s in the long outburst top light curves of HS 2325+8205. Moreover, we find that the oscillation intensity of the QPOs of HS 2325+8205 is related to the orbital phase, and the intensity in orbital phases 0.5-0.9 are stronger than in orbital phases 0.1-0.5. Therefore, the relationship between the oscillation intensity of QPOs and the orbital phase may become a research window for the origin of QPOs. In addition, we use the LSP to correct the orbital period of HS 2325+8205 as 0.19433475(6) d.

Due to the significant difference between the experimental measurements and the theoretical predictions of standard model (SM) for the value of $\mathcal{R}(D)$ of the semileptonic decay $B\to D\ell\bar{\nu}_{\ell}$, people speculate that it may be the evidence of new physics beyond the SM. Usually, the $D$-meson twist-2, 3 distribution amplitudes (DAs) $\phi_{2;D}(x,\mu)$, $\phi_{3;D}^p(x,\mu)$ and $\phi_{3;D}^\sigma(x,\mu)$ are the main error sources when using perturbative QCD factorization and light-cone QCD sum rules to study $B\to D\ell\bar{\nu}_{\ell}$. Therefore, it is important to get more reasonable and accurate behaviors for those DAs. Motivated by our previous work [Phys. Rev. D 104, no.1, 016021 (2021)] on pionic leading-twist DA, we revisit $D$-meson twist-2, 3 DAs $\phi_{2;D}(x,\mu)$, $\phi_{3;D}^p(x,\mu)$ and $\phi_{3;D}^\sigma(x,\mu)$. New sum rules formulae for the $\xi$-moments of these three DAs are suggested to obtain more accurate values. The light-cone harmonic oscillator models for those DAs are improved, and whose model parameters are determined by fitting the values of $\xi$-moments with the least squares method., Comment: 9 pages, 2 figures

In this paper, we calculate the scalar $a_0(980)$-meson leading-twist wavefunction by using light-cone harmonic oscillator model (LCHO). In which the model parameters are determined by fitting the $\xi$-moments $\langle\xi_{a_0}^n\rangle_\zeta$ of its light-cone distribution amplitudes. Then, the $a_0(980)$-meson leading-twist light-cone distribution amplitudes with three different scales $\zeta= (1.0, 2.0, 5.2)~{\rm GeV}$ are given. After constructing the relationship between $a_0(980)$-meson leading-twist parton distribution functions/valence quark distribution function and its LCHO wavefunction, we exhibit the $q^{a_0}(x,\zeta)$ and $x q^{a_0}(x,\zeta)$ with different scales. Furthermore, we also calculate the Mellin moments of the $a_0(980)$-meson's valence quark distribution function $\langle x^n q^{a_0}\rangle_\zeta$ with $n = (1,2,3)$, i.e. $\langle x q^{a_0}\rangle_{\zeta_5} = 0.026$, $\langle x^2 q^{a_0}\rangle_{\zeta_5} = 0.017$ and $\langle x^3 q^{a_0}\rangle_{\zeta_5} = 0.012$. Finally, the scale evolution for the ratio of the Mellin moments $x^n_{a_0}(\zeta,\zeta_k)$ are presented., Comment: 7 pages, 3 figures, comments welcome

Based on the scenario that the $K_0^\ast(1430)$ is viewed as the ground state of $s\bar{q}$ or $q\bar{s}$, we study the $K_0^\ast(1430)$ leading-twist distribution amplitude (DA) $\phi_{2;K_0^\ast}(x,\mu)$ with the QCD sum rules in the framework of background field theory. A more reasonable sum rule formula for $\xi$-moments $\langle\xi^n\rangle_{2;K_0^\ast}$ is suggested, which eliminates the influence brought by the fact that the sum rule of $\langle\xi^0_p\rangle_{3;K_0^\ast}$ cannot be normalized in whole Borel region. More accurate values of the first ten $\xi$-moments, $\langle\xi^n\rangle_{2;K_0^\ast} (n = 1,2,\cdots,10)$, are evaluated. A new light-cone harmonic oscillator (LCHO) model for $K_0^\ast(1430)$ leading-twist DA is established for the first times. By fitting the resulted values of $\langle\xi^n\rangle_{2;K_0^\ast} (n = 1,2,\cdots,10)$ via the least squares method, the behavior of $K_0^\ast(1430)$ leading-twist DA described with LCHO model is determined. Further, by adopting the light-cone QCD sum rules, we calculate the $B_s,D_s \to K_0^\ast(1430)$ transition form factors and branching fractions of the semileptonic decays $B_s,D_s \to K_0^\ast(1430) \ell \nu_\ell$. The corresponding numerical results can be used to extract the Cabibbo-Kobayashi-Maskawa matrix elements by combining the relative experimental data in the future., Comment: 16 pages, 4 figures

In this paper, moments of $a_0(980)$-meson twist-2 light-cone distribution amplitudes were deeply researched by using QCD sum rules approach within background field theory. Up to 9th-order accuracy, we present $\langle\xi_{2;a_0}^n\rangle|_{\mu_0}$ at the initial scale $\mu_0 = 1~{\rm GeV}$, i.e. $\langle\xi^1_{2;a_0}\rangle|_{\mu_0} = -0.307(43)$, $\langle\xi^3_{2;a_0}\rangle|_{\mu_0} = -0.181(34)$, $\langle\xi^5_{2;a_0}\rangle|_{\mu_0} = -0.078(28)$, $\langle\xi^7_{2;a_0}\rangle|_{\mu_0} = -0.049(26)$, $\langle\xi^9_{2;a_0}\rangle|_{\mu_0} = -0.036(24)$, respectively. An improved light-cone harmonic oscillator model for $a_0(980)$-meson twist-2 light-cone distribution amplitudes is adopted, where its parameters are fixed by using the least squares method based on the $\langle\xi_{2;a_0}^n\rangle|_{\mu_0}$, and their goodness of fit reach to $95.4\%$. Then, we calculate the $D\to a_0(980)$ transition form factors within the light-cone sum rules approach, and at largest recoil point, we obtain $f_+^{D\to a_0}(0) = 1.058^{+0.068}_{-0.035}$ and $f_-^{D\to a_0}(0) = 0.764^{+0.044}_{-0.036}$. As a further application, the branching fractions of the $D\to a_0(980)\ell\bar\nu_\ell$ semileptonic decays are given. Taking the decay $a_0(980)\to \eta\pi$ into consideration, we obtain ${\cal B}(D^0 \to a_0(980)^- (\to \eta \pi^-) e^+\nu_e) =(1.330^{+0.216}_{-0.134})\times10^{-4}$, ${\cal B}(D^+\to a_0(980)^0(\to \eta \pi^0)e^+\nu_e)=(1.675^{+0.272}_{-0.169})\times10^{-4}$, which are consistent with the BESIII collaboration and PDG data within errors. Finally, we present the angle observables of forward-backward asymmetries, $q^2$-differential flat terms and lepton polarization asymmetry of the semileptonic decay $D\to a_0(980)\ell\bar\nu_\ell$., Comment: 14 pages, 6 figures, more discussions added, accepted for publication in Nuclear Physics A

In this paper, we tackle the problem of domain shift. Most existing methods perform training on multiple source domains using a single model, and the same trained model is used on all unseen target domains. Such solutions are sub-optimal as each target domain exhibits its own specialty, which is not adapted. Furthermore, expecting single-model training to learn extensive knowledge from multiple source domains is counterintuitive. The model is more biased toward learning only domain-invariant features and may result in negative knowledge transfer. In this work, we propose a novel framework for unsupervised test-time adaptation, which is formulated as a knowledge distillation process to address domain shift. Specifically, we incorporate Mixture-of-Experts (MoE) as teachers, where each expert is separately trained on different source domains to maximize their specialty. Given a test-time target domain, a small set of unlabeled data is sampled to query the knowledge from MoE. As the source domains are correlated to the target domains, a transformer-based aggregator then combines the domain knowledge by examining the interconnection among them. The output is treated as a supervision signal to adapt a student prediction network toward the target domain. We further employ meta-learning to enforce the aggregator to distill positive knowledge and the student network to achieve fast adaptation. Extensive experiments demonstrate that the proposed method outperforms the state-of-the-art and validates the effectiveness of each proposed component. Our code is available at https://github.com/n3il666/Meta-DMoE., Comment: Accepted at NeurIPS2022

So far, the behavior of the pionic leading-twist distribution amplitude (DA) $\phi_{2;\pi}(x,\mu)$ $-$ which is universal physical quantity and enters the high-energy processes involving pion based on the factorization theorem $-$ has not been completely consistent. The form of $\phi_{2;\pi}(x,\mu)$ is usually described by phenomenological models and constrained by the experimental data of the exclusive processes containing pion or the moments calculated with the QCD sum rules and lattice QCD theory. Obviously, an appropriate model is very important for us to determine the exact behavior of $\phi_{2;\pi}(x,\mu)$. In this paper, by adopting the least squares method to fit the $\xi$-moments calculated with QCD sum rules based on the background field theory, we perform an analysis for several commonly used models of the pionic leading-twist DA in the literature, such as the truncation form of the Gegenbauer polynomial series, the light-cone harmonic oscillator model, the form from the Dyson-Schwinger equations, the model from the light-front holographic AdS/QCD and a simple power-law parametrization form., Comment: 10 pages, 2 figures

We demonstrate a kind of grating coupler that generates high quality flat-top beam with a small beamwidth from photonic integrated circuits into free-space. The grating coupler is designed on a silicon-on-insulator wafer with a 220 nm-thick silicon layer and consists of a dual-etch grating (DG) and a distributed Bragg reflector (DBR). By adjusting the structural parameters of DG and DBR, a pixel-level (6.6 um) flat-top beam with the vertical radiation of -0.5 dB and the mode match of 97% at 1550 nm is realized. Furthermore, a series of high-efficiency grating couplers are designed to create flat-top beam with different scales.

In the paper, we study the $B_s\to D_s$ transition form factors by using the light-cone sum rules within the framework of heavy quark effective field theory. We adopt a chiral current correlation function to do the calculation, the resultant transition form factors $f_+^{B_s\to D_s}(q^2)$ and $f_0^{B_s\to D_s}(q^2)$ are dominated by the contribution of $D_s$-meson leading-twist distribution amplitude, while the contributions from less certain $D_s$-meson twist-3 distribution amplitudes are greatly suppressed. At the largest recoil point, we obtain $f_{+,0}^{B_s \to D_s}(0)=0.533^{+0.112}_{-0.094}$. By further extrapolating the transition form factors into all the physically allowable $q^2$ region with the help of the $z$-series parametrization approach, we calculate the branching fractions $\mathcal{B}(B_s \to D_s \ell^\prime \nu_{\ell^\prime})$ with $(\ell^\prime= e,\mu)$ and $\mathcal{B}(B_s \to D_s \tau \nu_\tau)$, which gives $\mathcal{R}(D_s)=0.334\pm 0.017$., Comment: 9 pages, 7 figures

In this paper, we calculate the ratio of Cabibbo-Kobayashi-Maskawa matrix elements, $|V_{ub}|/|V_{cb}|$, based on the semileptonic decay $B_s^0\to K^-\mu^+\nu_\mu$. Its key component, the $B_s\to K$ transition form factor $f^{B_s\to K}_+(q^2)$, is studied within the QCD light-cone sum rules approach by using a chiral correlator. The derived $f^{B_s\to K}_+(q^2)$ is dominated by the leading-twist part, and to improve its precision, we construct a new model for the kaon leading-twist distribution amplitude $\phi_{2;K}(x,\mu)$, whose parameters are fixed by using the least squares method with the help of the moments calculated by using the QCD sum rules within the background field theory. The first four moments at the initial scale $\mu_0 = 1~{\rm GeV}$ are, $\langle \xi^1\rangle _{2;K} = -0.0438^{+0.0053}_{-0.0075}$, $\langle \xi^2\rangle _{2;K} = 0.262 \pm 0.010$, $\langle \xi^3\rangle _{2;K} = -0.0210^{+0.0024}_{-0.0035}$ and $\langle \xi^4\rangle _{2;K} = 0.132 \pm 0.006$, respectively. And their corresponding Gegenbauer moments are, $a^{2;K}_1 = -0.0731^{+0.0089}_{-0.0124}$, $a^{2;K}_2 = 0.182^{+0.029}_{-0.030}$, $a^{2;K}_3 = -0.0114^{+0.0008}_{-0.0016}$ and $a^{2;K}_4 = 0.041^{-0.003}_{+0.005}$, respectively. At the large recoil region, we obtain $f^{B_s\to K} _+ (0) = 0.270^{+0.022}_{-0.030}$. By extrapolating $f^{B_s\to K}_+(q^2)$ to all the physical allowable region, we obtain a $|V_{ub}|$-independent decay width for the semileptonic decay $B_s^0\to K^-\mu^+\nu_\mu$, $5.626^{+1.271}_{-0.864} \times 10^{-12}\ {\rm GeV}$, which then leads to $|V_{ub}|/|V_{cb}| = 0.072\pm0.005$., Comment: 34 pages, 11 figures, comments welcome

The DArk Matter Particle Explorer (DAMPE) is well suitable for searching for monochromatic and sharp $\gamma$-ray structures in the GeV$-$TeV range thanks to its unprecedented high energy resolution. In this work, we search for $\gamma$-ray line structures using five years of DAMPE data. To improve the sensitivity, we develop two types of dedicated data sets (including the BgoOnly data which is the first time to be used in the data analysis for the calorimeter-based gamma-ray observatories) and adopt the signal-to-noise ratio optimized regions of interest (ROIs) for different DM density profiles. No line signals or candidates are found between 10 and 300 GeV in the Galaxy. The constraints on the velocity-averaged cross section for $\chi\chi \to \gamma\gamma$ and the decay lifetime for $\chi \to \gamma\nu$, both at 95% confidence level, have been calculated and the systematic uncertainties have been taken into account. Comparing to the previous Fermi-LAT results, though DAMPE has an acceptance smaller by a factor of $\sim 10$, similar constraints on the DM parameters are achieved and below 100 GeV the lower limits on the decay lifetime are even stronger by a factor of a few. Our results demonstrate the potential of high-energy-resolution observations on dark matter detection., Comment: 14 pages, 8 figures. Update the content to keep up with the published version

In the paper, we investigate the moments $\langle\xi_{2;a_1}^{\|;n}\rangle$ of the axial-vector $a_1(1260)$-meson distribution amplitude by using the QCD sum rules approach under the background field theory. By considering the vacuum condensates up to dimension-six and the perturbative part up to next-to-leading order QCD corrections, its first five moments at an initial scale $\mu_0=1~{\rm GeV}$ are $\langle\xi_{2;a_1}^{\|;2}\rangle|_{\mu_0} = 0.223 \pm 0.029$, $\langle\xi_{2;a_1}^{\|;4}\rangle|_{\mu_0} = 0.098 \pm 0.008$, $\langle\xi_{2;a_1}^{\|;6}\rangle|_{\mu_0} = 0.056 \pm 0.006$, $\langle\xi_{2;a_1}^{\|;8}\rangle|_{\mu_0} = 0.039 \pm 0.004$ and $\langle\xi_{2;a_1}^{\|;10}\rangle|_{\mu_0} = 0.028 \pm 0.003$, respectively. We then construct a light-cone harmonic oscillator model for $a_1(1260)$-meson longitudinal twist-2 distribution amplitude $\phi_{2;a_1}^{\|}(x,\mu)$, whose model parameters are fitted by using the least squares method. As an application of $\phi_{2;a_1}^{\|}(x,\mu)$, we calculate the transition form factors (TFFs) of $D\to a_1(1260)$ in large and intermediate momentum transfers by using the QCD light-cone sum rules approach. At the largest recoil point ($q^2=0$), we obtain $ A(0) = 0.130_{ - 0.013}^{ + 0.015}$, $V_1(0) = 1.898_{-0.121}^{+0.128}$, $V_2(0) = 0.228_{-0.021}^{ + 0.020}$, and $V_0(0) = 0.217_{ - 0.025}^{ + 0.023}$. By applying the extrapolated TFFs to the semi-leptonic decay $D^{0(+)} \to a_1^{-(0)}(1260)\ell^+\nu_\ell$, we obtain ${\cal B}(D^0\to a_1^-(1260) e^+\nu_e) = (5.261_{-0.639}^{+0.745}) \times 10^{-5}$, ${\cal B}(D^+\to a_1^0(1260) e^+\nu_e) = (6.673_{-0.811}^{+0.947}) \times 10^{-5}$, ${\cal B}(D^0\to a_1^-(1260) \mu^+ \nu_\mu)=(4.732_{-0.590}^{+0.685}) \times 10^{-5}$, ${\cal B}(D^+ \to a_1^0(1260) \mu^+ \nu_\mu)=(6.002_{-0.748}^{+0.796}) \times 10^{-5}$., Comment: 13 pages, 4 figures, version published in EPJC