Search

Narrow-gap rare-earth Zintl phases frequently exhibit fascinating physical phenomena due to their various crystal structures, complex magnetic properties, and tunable transport behaviors. Here we report the synthesis, magnetic, thermodynamic, and transport properties of a Eu-containing Zintl arsenide, Eu$_{11}$Zn$_6$As$_{12}$, which consists of infinite chains of Eu cations and anionic frameworks constructed from corner-sharing ZnAs$_4$ tetrahedra. Eu$_{11}$Zn$_6$As$_{12}$ exhibits complicated magnetic behavior owing to intricate exchange interactions mediated by the discrete anionic fragments. Two long-range magnetic transitions at 22 K ($T_\mathrm{N}$) and 9 K ($T^*$), as well as exceptionally strong ferromagnetic fluctuations around 29 K ($T_\mathrm{F}$), are indicated by the susceptibility, heat capacity and resistivity measurements. Besides, Eu$_{11}$Zn$_6$As$_{12}$ displays metallic behavior, attributable to the hole carriers doped by slight Eu vacancies or the mixed valence of Eu$^{2+}$ and Eu$^{3+}$. A prominent resistivity peak occurs around $T_\mathrm{N}$, which is rapidly suppressed by the applied field, leading to a prominent negative magnetoresistance effect. A resistivity hysteresis is observed below 5 K, caused by a small net ferromagnetic component. Our study presents the distinct magnetic and transport properties of Eu$_{11}$Zn$_6$As$_{12}$, and further experiments are required to elucidate the origin of these novel behaviors. Moreover, our findings demonstrate that Eu-based Zintl phases are a fertile ground to study the interplay between magnetism and charge transport.

We study integrals appearing in one-loop amplitudes in string theory, and in particular their analytic continuation based on a string theoretic analog of the $i\varepsilon$-prescription of quantum field theory. For various zero- and two-point one-loop amplitudes of both open and closed strings, we prove that this analytic continuation is equivalent to a regularization using generalized exponential integrals. Our approach provides exact expressions in terms of the degeneracies at each mass level. For one-loop amplitudes with boundaries, our result takes the form of a linear combination of three partition functions at different temperatures depending on a variable $T_0$, yet their sum is independent of this variable. The imaginary part of the amplitudes can be read off in closed form, while the real part is amenable to numerical evaluation. While the expressions are rather different, we demonstrate agreement of our approach with the contour put forward by Eberhardt-Mizera (2023) following the Hardy-Ramanujan-Rademacher circle method., Comment: 33 pages + appendices

Recent experiments suggest a new paradigm towards novel colossal magnetoresistance (CMR) in a family of materials EuM$_2$X$_2$(M=Cd, In, Zn; X=P, As), distinct from the traditional avenues involving Kondo-RKKY crossovers, magnetic phase transitions with structural distortions, or topological phase transitions. Here, we use angle-resolved photoemission spectroscopy (ARPES) and density functional theory (DFT) calculations to explore their origin, particularly focusing on EuCd$_2$P$_2$. While the low-energy spectral weight royally tracks that of the resistivity anomaly near the temperature with maximum magnetoresistance (T$_{MR}$) as expected from transport-spectroscopy correspondence, the spectra are completely incoherent and strongly suppressed with no hint of a Landau quasiparticle. Using systematic material and temperature dependence investigation complemented by theory, we attribute this non-quasiparticle caricature to the strong presence of entangled magnetic and lattice interactions, a characteristic enabled by the $p$-$f$ mixing. Given the known presence of ferromagnetic clusters, this naturally points to the origin of CMR being the scattering of spin-polarized polarons at the boundaries of ferromagnetic clusters. These results are not only illuminating to investigate the strong correlations and topology in EuCd$_2$X$_2$ family, but, in a broader view, exemplify how multiple cooperative interactions can give rise to extraordinary behaviors in condensed matter systems.

As a bridge connecting the matching polynomial and the Laplacian matching polynomial of graphs, the subdivision method is expected to be useful for investigating the Laplacian matching polynomial. In this paper, we study applications of the method from three aspects. We prove that the zero sequence of the Laplacian matching polynomial of a graph majorizes its degree sequence, establishing a dual relation between the Laplacian matching polynomial and the characteristic polynomial of the signless Laplacian matrix of graphs. In addition, from different viewpoints, we give a new combinatorial interpretations for the coefficients of the Laplacian matching polynomial.

A longstanding goal of artificial general intelligence is highly capable generalists that can learn from diverse experiences and generalize to unseen tasks. The language and vision communities have seen remarkable progress toward this trend by scaling up transformer-based models trained on massive datasets, while reinforcement learning (RL) agents still suffer from poor generalization capacity under such paradigms. To tackle this challenge, we propose Meta Decision Transformer (Meta-DT), which leverages the sequential modeling ability of the transformer architecture and robust task representation learning via world model disentanglement to achieve efficient generalization in offline meta-RL. We pretrain a context-aware world model to learn a compact task representation, and inject it as a contextual condition to the causal transformer to guide task-oriented sequence generation. Then, we subtly utilize history trajectories generated by the meta-policy as a self-guided prompt to exploit the architectural inductive bias. We select the trajectory segment that yields the largest prediction error on the pretrained world model to construct the prompt, aiming to encode task-specific information complementary to the world model maximally. Notably, the proposed framework eliminates the requirement of any expert demonstration or domain knowledge at test time. Experimental results on MuJoCo and Meta-World benchmarks across various dataset types show that Meta-DT exhibits superior few and zero-shot generalization capacity compared to strong baselines while being more practical with fewer prerequisites. Our code is available at https://github.com/NJU-RL/Meta-DT., Comment: NeurIPS 2024. TLDR: We leverage the sequential modeling ability of the transformer architecture and robust task representation learning via world model disentanglement to achieve efficient generalization in offline meta-RL

This paper is concerned with the boundary-layer solutions of the singular Keller-Segel model proposed by Keller-Segel (1971) in a multi-dimensional domain, where the zero-flux boundary condition is imposed to the cell while inhomogeneous Dirichlet boundary condition to the nutrient. The steady-state problem of the Keller-Segel system is reduced to a scalar Dirichlet nonlocal elliptic problem with singularity. Studying this nonlocal problem, we obtain the unique steady-state solution which possesses a boundary spike-layer profile as nutrient diffusion coefficient $\varepsilon>0$ tends to zero. When the domain is radially symmetric, we find the explicit expansion for the slope of boundary-layer profiles at the boundary and boundary-layer thickness in terms of the radius as $\varepsilon>0$ is small, which pinpoints how the boundary curvature affects the boundary-layer profile and thickness. Furthermore, we establish the nonlinear exponential stability of the boundary-layer steady-state solution for the radially symmetric domain. The main challenge encountered in the analysis is that the singularity will arise when the nutrient diffusion coefficient $\varepsilon>0$ is small for both stationary and time-dependent problems. By relegating the nonlocal steady-state problem to local problems and performing a delicate analysis using the barrier method and Fermi coordinates, we can obtain refined estimates for the solution of local steady-state problem near the boundary. This strategy finally helps us to find the asymptotic profile of the solution to the nonlocal problem as $\varepsilon \to 0$ so that the singularity is accurately captured and hence properly handled to achieve our results., Comment: 8 figures

We investigate the supersymmetry breaking soft terms for all the viable models in the complete landscape of three-family supersymmetric Pati-Salam models arising from intersecting D6-branes on a $\mathbb{T}^6/(\mathbb{Z}_2\times \mathbb{Z}_2)$ orientifold in type IIA string theory. The calculations are performed in the general scenario of $u$-moduli dominance with the $s$-moduli turned on, where the soft terms remain independent of the Yukawa couplings and the Wilson lines. The results for the trilinear coupling, gaugino-masses, squared-mass parameters of squarks, sleptons and Higgs depend on the brane wrapping numbers and the susy breaking parameters. We find that unlike the Yukawa couplings which remain unchanged for the models dual under the exchange of two SU(2) sectors, the corresponding soft term parameters only match for the trilinear coupling and the mass of the gluino. This can be explained by the internal geometry where the Yukawa interactions depend only on the triangular areas of the worldsheet instantons while the soft terms have an additional dependence on the orientation-angles of D6-branes in the three two-tori. In the special limit of parameter space we find universal masses for the Higgs and the gauginos., Comment: 61 pages + appendix, 18 figures. arXiv admin note: substantial text overlap with arXiv:2409.09110

Remote sensing image plays an irreplaceable role in fields such as agriculture, water resources, military, and disaster relief. Pixel-level interpretation is a critical aspect of remote sensing image applications; however, a prevalent limitation remains the need for extensive manual annotation. For this, we try to introduce open-vocabulary semantic segmentation (OVSS) into the remote sensing context. However, due to the sensitivity of remote sensing images to low-resolution features, distorted target shapes and ill-fitting boundaries are exhibited in the prediction mask. To tackle this issue, we propose a simple and general upsampler, SimFeatUp, to restore lost spatial information in deep features in a training-free style. Further, based on the observation of the abnormal response of local patch tokens to [CLS] token in CLIP, we propose to execute a straightforward subtraction operation to alleviate the global bias in patch tokens. Extensive experiments are conducted on 17 remote sensing datasets spanning semantic segmentation, building extraction, road detection, and flood detection tasks. Our method achieves an average of 5.8%, 8.2%, 4.0%, and 15.3% improvement over state-of-the-art methods on 4 tasks. All codes are released. \url{https://earth-insights.github.io/SegEarth-OV}

In this work, we introduce explicit P-waves to construct the diquark operators, then construct the local four-quark currents to explore the hidden-charm tetraquark states with the $J^{PC}=0^{++}$, $1^{+-}$ and $2^{++}$ in the framework of the QCD sum rules at length. Our calculations indicate tiny light-flavor $SU(3)$ breaking effects on the tetraquark masses due to the special currents and the predictions support assigning the $X(4775)$ and $X(4500)$ as the $[uc]_{\widehat{V}}[\overline{uc}]_{\widehat{V}}-[dc]_{\widehat{V}}[\overline{dc}]_{\widehat{V}}$ and $[sc]_{\widehat{V}}[\overline{sc}]_{\widehat{V}}$ tetraquark states with the $J^{PC}=0^{++}$ respectively, and assigning the $Z_{c}(4600)$ and $Z_{\bar{c}\bar{s}}(4600)$ as the $[uc]_{\widehat{V}}[\overline{dc}]_{\widehat{V}}$ and $[qc]_{\widehat{V}}[\overline{sc}]_{\widehat{V}}$ tetraquark states with the $J^{PC}=1^{+-}$ respectively, and thus account for the LHCb's data., Comment: 12 pages, 2 figures

We examine the vacuum stability of gauge symmetry breaking in five dimensions, compactified on the $S_1/(\mathbb{Z}_2 \times \mathbb{Z}'_2)$ orbifold. We consider $SU(N)$, $Sp(N)$, $SO(2N)$ and $SO(2N+1)$ theories in the bulk, and provide an exhaustive classification of possible parity assignments that lead to stable orbifolds and of the corresponding symmetry breaking patterns. We use these results in the search for viable asymptotic grand unification theories (aGUT), testing the stability criteria on models based on $SU(6)$ and $SU(8)$. As a result, we identify two viable aGUTs: a unique $SU(6)$ pathway down to the Standard Model, and one $SU(8)$ model leading to an intermediate Pati-Salam partial unification., Comment: 44 pages, 4 figures

3D Gaussian Splatting demonstrates excellent quality and speed in novel view synthesis. Nevertheless, the huge file size of the 3D Gaussians presents challenges for transmission and storage. Current works design compact models to replace the substantial volume and attributes of 3D Gaussians, along with intensive training to distill information. These endeavors demand considerable training time, presenting formidable hurdles for practical deployment. To this end, we propose MesonGS, a codec for post-training compression of 3D Gaussians. Initially, we introduce a measurement criterion that considers both view-dependent and view-independent factors to assess the impact of each Gaussian point on the rendering output, enabling the removal of insignificant points. Subsequently, we decrease the entropy of attributes through two transformations that complement subsequent entropy coding techniques to enhance the file compression rate. More specifically, we first replace rotation quaternions with Euler angles; then, we apply region adaptive hierarchical transform to key attributes to reduce entropy. Lastly, we adopt finer-grained quantization to avoid excessive information loss. Moreover, a well-crafted finetune scheme is devised to restore quality. Extensive experiments demonstrate that MesonGS significantly reduces the size of 3D Gaussians while preserving competitive quality., Comment: 18 pages, 8 figures, ECCV 2024

Recently, the complete landscape of three-family supersymmetric Pati-Salam models from intersecting D6-branes on a type IIA $\mathbb{T}^6/(\mathbb{Z}_2\times \mathbb{Z}_2)$ orientifold has been enumerated consisting of 33 independent models with distinct gauge coupling relations at the string scale. Here, we study the phenomenology of all such models by providing the detailed particle spectra and the analysis of the possible 3-point and the 4-point Yukawa interactions in order to accommodate all standard-model fermion masses and mixings. We find that only 17 models contain viable Yukawa textures to explain quarks masses, charged-leptons' masses, neutrino-masses, quarks' mixings and leptons' mixings. These viable models split into four classes, viz. a single model with 3 Higgs fields from the bulk and sixteen models with either 6, 9 or 12 Higgs from the $\mathcal{N}=2$ sector. The models perform successively better with the increasing number of Higgs pairs. Remarkably, the class of models with 12 Higgs naturally predicts the Dirac-type neutrino masses in normal ordering consistent with both the experimental constraints as well as the bounds from the swampland program., Comment: 89 pages + appendix, 21 figures, accepted in JHEP

Neural Radiance Fields have achieved success in creating powerful 3D media representations with their exceptional reconstruction capabilities. However, the computational demands of volume rendering pose significant challenges during model training. Existing acceleration techniques often involve redesigning the model architecture, leading to limitations in compatibility across different frameworks. Furthermore, these methods tend to overlook the substantial memory costs incurred. In response to these challenges, we introduce an expansive supervision mechanism that efficiently balances computational load, rendering quality and flexibility for neural radiance field training. This mechanism operates by selectively rendering a small but crucial subset of pixels and expanding their values to estimate the error across the entire area for each iteration. Compare to conventional supervision, our method effectively bypasses redundant rendering processes, resulting in notable reductions in both time and memory consumption. Experimental results demonstrate that integrating expansive supervision within existing state-of-the-art acceleration frameworks can achieve 69% memory savings and 42% time savings, with negligible compromise in visual quality., Comment: 12 pages, 7 figures

We systematically study the 1D charmonium spin-triplet (with the $J^{PC}=1^{--}, 2^{--}, 3^{--}$) and spin-singlet (with the $J^{PC}=2^{-+}$) via the QCD sum rules in comparison with the present experimental results. More experimental data on the D-wave charmonium states will help us to unravel the mass spectrum of the charmonium states near the open-charm thresholds., Comment: 11 pages, 27 figures

We introduce a relative P-wave to construct the vector doubly-charm diquark $(\widetilde{V})$ therefore the scalar and tensor tetraquark currents to investigate the decay widths of the fully-charm tetraquark states with the $J^{PC}=0^{++}$, $1^{+-}$ and $2^{++}$ via the QCD sum rules. We observe that the total width of the ground state $\widetilde{V}\overline{\widetilde{V}}$-type scalar tetraquark state is compatible with that of the $X(6552)$ within the range of uncertainties, and the branching ratios are quite different from that of the first radial excitation of the $A\bar{A}$-type scalar tetraquark state. Other predictions can be verified in the future experiments to shed light on the nature of the fully-charm tetraquark states., Comment: 12 pages, 2 figures

Over the past five years, significant progress has been made in understanding the magnetism and electronic properties of CaAl$_2$Si$_2$-type Eu$M_2X_2$ ($M$ = Zn, Cd; $X$ = P, As) compounds. Prior theoretical work and experimental studies suggested that EuCd$_2$As$_2$ had the potential to host rich topological phases, particularly an ideal magnetic Weyl semimetal state when the spins are polarized along the c axis. However, this perspective is challenged by recent experiments utilizing samples featuring ultra-low carrier densities, as well as meticulous calculations employing various approaches. Nonetheless, the Eu$M_2X_2$ family still exhibit numerous novel properties that remain to be satisfactorily explained, such as the giant nonlinear anomalous Hall effect and the colossal magnetoresistance effect. Moreover, Eu$M_2X_2$ compounds can be transformed from semiconducting antiferromagnets to metallic ferromagnets by introducing a small number of carriers or applying external pressure, and a further increase in the ferromagnetic transition temperature can be achieved by reducing the unit cell volume. These features make the Eu$M_2X_2$ family a fertile platform for studying the interplay between magnetism and charge transport, and an excellent candidate for applications in spintronics. This paper presents a comprehensive review of the magnetic and transport behaviors of Eu$M_2X_2$ compounds with varying carrier densities, as well as the current insights into these characteristics. An outlook for future research opportunities is also provided.

We propose a novel class of Dark Matter (DM) candidates in the form of a heavy composite Axion-Like Particle (ALP) with highly suppressed electromagnetic interactions, being stable even for masses exceeding the GeV scale. We argue that such a composite ALP emerges as a bound state -- the dark glueball -- due to confinement in a pure Yang-Mills dark sector. In a minimal ultraviolet complete QCD-like model, cosmological production of dark gluons as well as photons occurs via heavy fermion annihilation which effectively reheats both the dark and visible sectors setting up their temperature scales. Furthermore, effective interactions between glueballs and photons, resembling those of standard ALPs, are radiatively generated by heavy fermion loops. Consequently, DM glueballs interacting with photons are dubbed `Glueball ALPs' (GALPs). We uncover novel phenomenology of GALPs focusing on their unique astrophysical and cosmological signatures., Comment: 7 pages, 1 figure

We present proper genuine multipartite entanglement (GME) measures for arbitrary multipartite and dimensional systems. By using the volume of concurrence regular polygonal pyramid we first derive the GME measure of four-partite quantum systems. From our measure it is verified that the GHZ state is more entangled than the W state. Then we study the GME measure for multipartite quantum states in arbitrary dimensions. A well defined GME measure is constructed based on the volume of the concurrence regular polygonal pyramid. Detailed example shows that our measure can characterize better the genuine multipartite entanglements.

Inspired by the great progress in the observations of charmonium-like states in recent years, we perform a systematic analysis about the ground states and the first radially excited states of $qc\bar{q}\bar{c}$ ($q$=$u/d$ and $s$) tetraquark systems. Their mass spectra, root mean square (r.m.s.) radii and radial density distributions are predicted within the framework of relativized quark model. By comparing with experimental data, some potential candidates for hidden-charm tetraquark states are suggested. For $qc\bar{q}\bar{c}$ ($q$=$u/d$) system, if $Z_{c}(3900)$ is supposed to be a compact tetraquark state with $J^{PC}=1^{+-}$, $Z(4430)$ can be interpreted as the first radially excited states of $Z_{c}(3900)$. Another broad structure $Z_{c}(4200)$ can also be explained as a partner of $Z_{c}(3900)$, and it arise from a higher state with $J^{PC}=1^{+-}$. In addition, theoretical predictions indicate that the possible assignments for $X(3930)$, $X(4050)$ and $X(4250)$ are low lying $0^{++}$ tetraquark states. As for the $sc\bar{s}\bar{c}$ system, $X(4140)$ and $X(4274)$ structures can be interpreted as this type of tetraquark states with $J^{PC}=1^{++}$, and $X(4350)$ can be described as a $sc\bar{s}\bar{c}$ tetraquark with $J^{PC}=0^{++}$. With regard to $qc\bar{s}\bar{c}$ ($q$=$u/d$) system, we find two potential candidates for this type of tetraquark, which are $Z_{cs}(4000)$ and $Z_{cs}(4220)$ structures. The measured masses of these two structures are in agreement with theoretical predictions for the $1^{+}$ state.

A force field is a critical component in molecular dynamics simulations for computational drug discovery. It must achieve high accuracy within the constraints of molecular mechanics' (MM) limited functional forms, which offers high computational efficiency. With the rapid expansion of synthetically accessible chemical space, traditional look-up table approaches face significant challenges. In this study, we address this issue using a modern data-driven approach, developing ByteFF, an Amber-compatible force field for drug-like molecules. To create ByteFF, we generated an expansive and highly diverse molecular dataset at the B3LYP-D3(BJ)/DZVP level of theory. This dataset includes 2.4 million optimized molecular fragment geometries with analytical Hessian matrices, along with 3.2 million torsion profiles. We then trained an edge-augmented, symmetry-preserving molecular graph neural network (GNN) on this dataset, employing a carefully optimized training strategy. Our model predicts all bonded and non-bonded MM force field parameters for drug-like molecules simultaneously across a broad chemical space. ByteFF demonstrates state-of-the-art performance on various benchmark datasets, excelling in predicting relaxed geometries, torsional energy profiles, and conformational energies and forces. Its exceptional accuracy and expansive chemical space coverage make ByteFF a valuable tool for multiple stages of computational drug discovery., Comment: ByteFF, a machine learning parametrized MMFF. Code available at https://github.com/bytedance/byteff

The uncertainty relation reveals the intrinsic difference between the classical world and the quantum world. We investigate the quantum uncertainty relation of quantum channel in qubit systems. Under two general measurement bases, we first derive the quantum uncertainty relation for quantum channels with respect to the relative entropy of coherence. Then we obtain the quantum uncertainty relation for unitary channels with respect to the $l_1$ norm of coherence. Some examples are given in detail.

A significant limitation of wide-bandgap materials is their low hole mobility related to localized holes with heavy effective masses ($m_h^*$). We identify in low-symmetric wide-bandgap compounds an anion-anion antibonding coupling (AAAC) effect as the intrinsic factor behind hole localization, which explains the extremely heavy $m_h^*$ and self-trapped hole (STH) formation observed in gallium oxide ($\beta$-$Ga_{2}O_{3}$). We propose a design principle for achieving light holes by manipulating AAAC, demonstrating that specific strain conditions can reduce $m_h^*$ in $\beta$-$Ga_{2}O_{3}$ from 4.77 $m_0$ to 0.38 $m_0$, making it comparable to the electron mass (0.28 $m_0$), while also suppressing STH. The light holes show significant anisotropy, potentially enabling two-dimensional transport in bulk material. This study provides a fundamental understanding of hole mass enhancement and STH formation in novel wide-bandgap materials and suggest new pathways for engineering hole mobilities., Comment: 22 pages, 1 table, 5 figures

Performing unsupervised domain adaptation on resource-constrained edge devices is challenging. Existing research typically adopts architecture optimization (e.g., designing slimmable networks) but requires expensive training costs. Moreover, it does not consider the considerable precision redundancy of parameters and activations. To address these limitations, we propose efficient unsupervised domain adaptation with ReTraining-Free Quantization (RTF-Q). Our approach uses low-precision quantization architectures with varying computational costs, adapting to devices with dynamic computation budgets. We subtly configure subnet dimensions and leverage weight-sharing to optimize multiple architectures within a single set of weights, enabling the use of pre-trained models from open-source repositories. Additionally, we introduce multi-bitwidth joint training and the SandwichQ rule, both of which are effective in handling multiple quantization bit-widths across subnets. Experimental results demonstrate that our network achieves competitive accuracy with state-of-the-art methods across three benchmarks while significantly reducing memory and computational costs.

A service-level objective (SLO) is a target performance metric of service that cloud vendors aim to ensure. Delivering optimized SLOs can enhance user satisfaction and improve the competitiveness of cloud vendors. As large language models (LLMs) are gaining increasing popularity across various fields, it is of great significance to optimize SLOs for LLM inference services. In this paper, we observe that adjusting the parameters of LLM inference engines can improve service performance, and the optimal parameter configurations of different services are different. Therefore, we propose SCOOT, an automatic performance tuning system to optimize SLOs for each LLM inference service by tuning the parameters of the inference engine. We first propose a generalized formulation of the tuning problem to handle various objectives and constraints between parameters, and SCOOT exploits the Bayesian optimization (BO) technique to resolve the problem via exploration and exploitation. Moreover, SCOOT adopts a random forest to learn hidden constraints during the tuning process to mitigate invalid exploration. To improve the tuning efficiency, SCOOT utilizes the parallel suggestion to accelerate the tuning process. Extensive experiments demonstrate that SCOOT can significantly outperform existing tuning techniques in SLO optimization while greatly improving the tuning efficiency.

We study the dynamics of the quantum battery capacity for the Bell-diagonal states under Markovian channels on the first subsystem. We show that the capacity increases for special Bell-diagonal states under amplitude damping channel. The sudden death of the capacity occurs under depolarizing channel. We also investigate the capacity evolution of Bell-diagonal states under Markovian channels on the first subsystem $n$ times. It is shown that the capacity under depolarizing channel decreases initially, then increases for small $n$ and tend to zero for large $n$. We find that under bit flip channel and amplitude damping channel, the quantum battery capacity of special Bell-diagonal states tends to a constant for large $n$, namely, the frozen capacity occurs. The dynamics of the capacity of the Bell-diagonal states under two independent same type local Markovian channels is also studied., Comment: 12 pages, 6 figures

For on-policy reinforcement learning, discretizing action space for continuous control can easily express multiple modes and is straightforward to optimize. However, without considering the inherent ordering between the discrete atomic actions, the explosion in the number of discrete actions can possess undesired properties and induce a higher variance for the policy gradient estimator. In this paper, we introduce a straightforward architecture that addresses this issue by constraining the discrete policy to be unimodal using Poisson probability distributions. This unimodal architecture can better leverage the continuity in the underlying continuous action space using explicit unimodal probability distributions. We conduct extensive experiments to show that the discrete policy with the unimodal probability distribution provides significantly faster convergence and higher performance for on-policy reinforcement learning algorithms in challenging control tasks, especially in highly complex tasks such as Humanoid. We provide theoretical analysis on the variance of the policy gradient estimator, which suggests that our attentively designed unimodal discrete policy can retain a lower variance and yield a stable learning process., Comment: IEEE Transactions on Neural Networks and Learning Systems

Analyzing the supersymmetric Pati-Salam landscape on a $\mathbb{T}^6/(\mathbb{Z}_2 \times \mathbb{Z}_2)$ orientifold in IIA string theory, we have found only two models that accurately account for all standard model fermion masses and mixings. The models are dual to each other under the exchange of two SU(2) sectors and feature 12 adjoint scalars, the maximum number allowed in the landscape, whose linear combination yields the two light Higgs eigenstates. Dirac neutrino-masses in normal ordering $(50.4,~10.5,~6.1)~\mathrm{meV}$ satisfying the swampland constraints are predicted, a testable prospect for string phenomenology., Comment: 5 pages

Instigated by the first $T_{cc}$ observation at the LHCb Collaboration, the spectroscopic properties of the entire isoscalar and isovector $T_{cc}$ family are systematically unveiled by means of multiple sorts of relativized and nonrelativistic diquark formalisms, encompassing the Godfrey-Isgur relativized diquark model, the modified Godfrey-Isgur relativized diquark model incorporating the color screening effects, the nonrelativistic diquark model with the Gaussian type hyperfine potential, and the nonrelativistic diquark model with the Yukawa type hyperfine potential. The theoretical outcomes of various diquark-antidiquark scenarios are inclined to categorize the $T_{cc}(3875)^+$ structure as the exemplary candidate of the $1S$-wave isoscalar axial-vector double-charm tetraquark state. In light of the diquark-antidiquark configuration, this work investigates the mixing angles of the orbitally excited isovector $T_{cc}$ states and the magic mixing angles of the ideal heavy-light tetraquarks for the first time. As the advancement of the experimental detection capability, these phenomenological prognostications will effectively boost the hunting for the prospective low-lying $T_{cc}$ states in the future., Comment: 23 pages

We present Open-CD, a change detection toolbox that contains a rich set of change detection methods as well as related components and modules. The toolbox started from a series of open source general vision task tools, including OpenMMLab Toolkits, PyTorch Image Models, etc. It gradually evolves into a unified platform that covers many popular change detection methods and contemporary modules. It not only includes training and inference codes, but also provides some useful scripts for data analysis. We believe this toolbox is by far the most complete change detection toolbox. In this report, we introduce the various features, supported methods and applications of Open-CD. In addition, we also conduct a benchmarking study on different methods and components. We wish that the toolbox and benchmark could serve the growing research community by providing a flexible toolkit to reimplement existing methods and develop their own new change detectors. Code and models are available at \url{https://github.com/likyoo/open-cd}. Pioneeringly, this report also includes brief descriptions of the algorithms supported in Open-CD, mainly contributed by their authors. We sincerely encourage researchers in this field to participate in this project and work together to create a more open community. This toolkit and report will be kept updated., Comment: 9 pages

Monogamy of entanglement is the fundamental property of quantum systems. By using two new entanglement measures based on dual entropy, the $S^{t}$-entropy entanglement and $T^{t}_q$-entropy entanglement measures, we present the general monogamy relations in multi-qubit quantum systems. We show that these newly derived monogamy inequalities are tighter than the existing ones. Based on these general monogamy relations, we construct the set of multipartite entanglement indicators for $N$-qubit states, which are shown to work well even for the cases that the usual concurrence-based indicators do not work. Detailed examples are presented to illustrate our results., Comment: 10 pages, 7 figures

In the present work, we construct the diquark-antidiquark type four-quark currents to investigate the mass spectrum of the ground state hidden-charm-hidden-strange tetraquark states with the quantum numbers $J^{PC}=0^{++}$, $1^{+-}$, $1^{++}$ and $2^{++}$ via the traditional QCD sum rules in a comprehensive way. We update old calculations, perform new calculations and analysis in a rigorous way, and take account of the net light-flavor $SU(3)$ breaking effects in a consistent way. And we make more reasonable identifications for the $X(3960)$, $X(4140)$, $X(4274)$, $X(4500)$, $X(4685)$ and $X(4700)$ and supersede some old identifications. Furthermore, we consider our previous theoretical predictions, and make reasonable/suitable identifications of the new LHCb states $h_c(4000)$ and $\chi_{c1}(4010)$., Comment: 21 pages, 1 figure

Asynchronous Federated Learning (AFL) confronts inherent challenges arising from the heterogeneity of devices (e.g., their computation capacities) and low-bandwidth environments, both potentially causing stale model updates (e.g., local gradients) for global aggregation. Traditional approaches mitigating the staleness of updates typically focus on either adjusting the local updating or gradient compression, but not both. Recognizing this gap, we introduce a novel approach that synergizes local updating with gradient compression. Our research begins by examining the interplay between local updating frequency and gradient compression rate, and their collective impact on convergence speed. The theoretical upper bound shows that the local updating frequency and gradient compression rate of each device are jointly determined by its computing power, communication capabilities and other factors. Building on this foundation, we propose an AFL framework called FedLuck that adaptively optimizes both local update frequency and gradient compression rates. Experiments on image classification and speech recognization show that FedLuck reduces communication consumption by 56% and training time by 55% on average, achieving competitive performance in heterogeneous and low-bandwidth scenarios compared to the baselines.

Active polymers possess numerous unique properties that are quite different from those observed in the system of small active molecule due to the intricate interplay between their activity and topological constraints. This study focuses on the conformational changes induced by activity, impacting effective stiffness and crucially influencing entanglement and dynamics. When the two terminals of a linear chain undergo active modification through coupling to a high-temperature thermal bath, there is a substantial increase in chain size, indicating a notable enhancement in effective stiffness. Unlike in passive semiflexible chains where stiffness predominantly affects local bond angles, activity-induced stiffness manifests at the scale of tens of monomers. While activity raises the ambient temperature, it significantly decreases diffusion by over an order of magnitude. The slowdown of dynamics observed can be attributed to increased entanglement due to chain elongation., Comment: 10 pages, 5 figures

We introduce PRANCE, a Vision Transformer compression framework that jointly optimizes the activated channels and reduces tokens, based on the characteristics of inputs. Specifically, PRANCE~ leverages adaptive token optimization strategies for a certain computational budget, aiming to accelerate ViTs' inference from a unified data and architectural perspective. However, the joint framework poses challenges to both architectural and decision-making aspects. Firstly, while ViTs inherently support variable-token inference, they do not facilitate dynamic computations for variable channels. To overcome this limitation, we propose a meta-network using weight-sharing techniques to support arbitrary channels of the Multi-head Self-Attention and Multi-layer Perceptron layers, serving as a foundational model for architectural decision-making. Second, simultaneously optimizing the structure of the meta-network and input data constitutes a combinatorial optimization problem with an extremely large decision space, reaching up to around $10^{14}$, making supervised learning infeasible. To this end, we design a lightweight selector employing Proximal Policy Optimization for efficient decision-making. Furthermore, we introduce a novel "Result-to-Go" training mechanism that models ViTs' inference process as a Markov decision process, significantly reducing action space and mitigating delayed-reward issues during training. Extensive experiments demonstrate the effectiveness of PRANCE~ in reducing FLOPs by approximately 50\%, retaining only about 10\% of tokens while achieving lossless Top-1 accuracy. Additionally, our framework is shown to be compatible with various token optimization techniques such as pruning, merging, and sequential pruning-merging strategies. The code is available at \href{https://github.com/ChildTang/PRANCE}{https://github.com/ChildTang/PRANCE}.

Polymer chains immersed in different solvent molecules exhibit diverse properties due to multiple spatiotemporal scales and complex interactions. Using molecular dynamics simulations, we study the conformational and static properties of tagged chains in different solvent molecules. Two types of solvent molecules were examined: one type consisted of chain molecules connected by bonds, while the other type consisted of individual bead molecules without any bonds. The only difference between the two solvent molecules lay in the chain connectivity. Our results show a compression of the tagged chains with the addition of bead or chain molecules. Chain molecule confinement induces a stronger compression compared to bead molecule confinement. In chain solvent molecules, the tagged chain's radius of gyration reached a minimum at a monomer volume fraction of $\sim0.3$. Notably, the probability distributions of chain size remain unchanged at different solvent densities, irrespective of whether the solvent consists of beads or polymers. Furthermore, as solvent density increases, a crossover from a unimodal to a bimodal distribution of bond angles is observed, indicating the presence of both compressed and expanded regions within the chain. The effective monomer-solvent interaction is obtained by calculating the partial radial distribution function and the potential of the mean force. In chain solvent, the correlation hole effect results in a reduced number of nearest neighbors around tagged monomers compared to bead solvents. The calculation of pore size distribution reveals that the solvent nonhomogeneity induced by chain connectivity leads to a broader distribution of pore sizes and larger pore dimensions at low volume fractions. These findings provide a deeper understanding of the conformational behavior of polymer chains in different solvent environments., Comment: 10 pages, 8 figures

Heterogeneous treatment effect (HTE) estimation is vital for understanding the change of treatment effect across individuals or subgroups. Most existing HTE estimation methods focus on addressing selection bias induced by imbalanced distributions of confounders between treated and control units, but ignore distribution shifts across populations. Thereby, their applicability has been limited to the in-distribution (ID) population, which shares a similar distribution with the training dataset. In real-world applications, where population distributions are subject to continuous changes, there is an urgent need for stable HTE estimation across out-of-distribution (OOD) populations, which, however, remains an open problem. As pioneers in resolving this problem, we propose a novel Stable Balanced Representation Learning with Hierarchical-Attention Paradigm (SBRL-HAP) framework, which consists of 1) Balancing Regularizer for eliminating selection bias, 2) Independence Regularizer for addressing the distribution shift issue, 3) Hierarchical-Attention Paradigm for coordination between balance and independence. In this way, SBRL-HAP regresses counterfactual outcomes using ID data, while ensuring the resulting HTE estimation can be successfully generalized to out-of-distribution scenarios, thereby enhancing the model's applicability in real-world settings. Extensive experiments conducted on synthetic and real-world datasets demonstrate the effectiveness of our SBRL-HAP in achieving stable HTE estimation across OOD populations, with an average 10% reduction in the error metric PEHE and 11% decrease in the ATE bias, compared to the SOTA methods., Comment: Accepted by ICDE'2024

Recent advancements in diffusion models, particularly the trend of architectural transformation from UNet-based Diffusion to Diffusion Transformer (DiT), have significantly improved the quality and scalability of image synthesis. Despite the incredible generative quality, the large computational requirements of these large-scale models significantly hinder the deployments in real-world scenarios. Post-training Quantization (PTQ) offers a promising solution by compressing model sizes and speeding up inference for the pretrained models while eliminating model retraining. However, we have observed the existing PTQ frameworks exclusively designed for both ViT and conventional Diffusion models fall into biased quantization and result in remarkable performance degradation. In this paper, we find that the DiTs typically exhibit considerable variance in terms of both weight and activation, which easily runs out of the limited numerical representations. To address this issue, we devise Q-DiT, which seamlessly integrates three techniques: fine-grained quantization to manage substantial variance across input channels of weights and activations, an automatic search strategy to optimize the quantization granularity and mitigate redundancies, and dynamic activation quantization to capture the activation changes across timesteps. Extensive experiments on the ImageNet dataset demonstrate the effectiveness of the proposed Q-DiT. Specifically, when quantizing DiT-XL/2 to W8A8 on ImageNet 256x256, Q-DiT achieves a remarkable reduction in FID by 1.26 compared to the baseline. Under a W4A8 setting, it maintains high fidelity in image generation, showcasing only a marginal increase in FID and setting a new benchmark for efficient, high-quality quantization in diffusion transformers. Code is available at \href{https://github.com/Juanerx/Q-DiT}{https://github.com/Juanerx/Q-DiT}.

Large language models (LLMs) iteratively generate text token by token, with memory usage increasing with the length of generated token sequences. The unpredictability of generation lengths makes it difficult to estimate the time and memory needed to process requests, posing a challenge for effective request scheduling. Conventional sequence-level scheduling (SLS) serves requests in a first-come first-served (FCFS) manner with static batching where requests with short generation lengths are delayed until those with long ones have finished generation, which hurts computational efficiency. Besides, to avoid out-of-memory (OOM) errors, SLS batches requests with a small batch size, which limits throughput. Recently proposed iteration-level scheduling (ILS) enhances computational efficiency with continuous batching to return completed requests timely and dynamically add new requests for processing. However, many ILS schedulers limit the number of parallel-processing requests to avoid OOM errors while achieving a fast inference speed, which compromises throughput. Moreover, existing SLS and ILS schedulers fail to balance the workload across multiple deployed LLM instances. To tackle these challenges, we propose slice-level scheduling (SCLS). By splitting the predefined maximal generation length limit into slices and serving batches slice by slice, it provides a precise range of serving time and memory usage for batched requests, laying the foundation for effective scheduling. Experiments confirm that compared with SLS and ILS schedulers, SCLS can improve throughput by up to 315.8% and greatly mitigate load imbalance with proposed batching and offloading algorithms., Comment: 13 pages, 22 figures

Detecting objects in real-world scenes is a complex task due to various challenges, including the vast range of object categories, and potential encounters with previously unknown or unseen objects. The challenges necessitate the development of public benchmarks and challenges to advance the field of object detection. Inspired by the success of previous COCO and LVIS Challenges, we organize the V3Det Challenge 2024 in conjunction with the 4th Open World Vision Workshop: Visual Perception via Learning in an Open World (VPLOW) at CVPR 2024, Seattle, US. This challenge aims to push the boundaries of object detection research and encourage innovation in this field. The V3Det Challenge 2024 consists of two tracks: 1) Vast Vocabulary Object Detection: This track focuses on detecting objects from a large set of 13204 categories, testing the detection algorithm's ability to recognize and locate diverse objects. 2) Open Vocabulary Object Detection: This track goes a step further, requiring algorithms to detect objects from an open set of categories, including unknown objects. In the following sections, we will provide a comprehensive summary and analysis of the solutions submitted by participants. By analyzing the methods and solutions presented, we aim to inspire future research directions in vast vocabulary and open-vocabulary object detection, driving progress in this field. Challenge homepage: https://v3det.openxlab.org.cn/challenge

Large language models (LLMs) have exhibited exciting progress in multiple scenarios, while the huge computational demands hinder their deployments in lots of real-world applications. As an effective means to reduce memory footprint and inference cost, quantization also faces challenges in performance degradation at low bit-widths. Understanding the impact of quantization on LLM capabilities, especially the generalization ability, is crucial. However, the community's main focus remains on the algorithms and models of quantization, with insufficient attention given to whether the quantized models can retain the strong generalization abilities of LLMs. In this work, we fill this gap by providing a comprehensive benchmark suite for this research topic, including an evaluation system, detailed analyses, and a general toolbox. Specifically, based on the dominant pipeline in LLM quantization, we primarily explore the impact of calibration data distribution on the generalization of quantized LLMs and conduct the benchmark using more than 40 datasets within two main scenarios. Based on this benchmark, we conduct extensive experiments with two well-known LLMs (English and Chinese) and four quantization algorithms to investigate this topic in-depth, yielding several counter-intuitive and valuable findings, e.g., models quantized using a calibration set with the same distribution as the test data are not necessarily optimal. Besides, to facilitate future research, we also release a modular-designed toolbox, which decouples the overall pipeline into several separate components, e.g., base LLM module, dataset module, quantizer module, etc. and allows subsequent researchers to easily assemble their methods through a simple configuration. Our benchmark suite is publicly available at https://github.com/TsingmaoAI/MI-optimize

Learning-based intelligent energy management systems for plug-in hybrid electric vehicles (PHEVs) are crucial for achieving efficient energy utilization. However, their application faces system reliability challenges in the real world, which prevents widespread acceptance by original equipment manufacturers (OEMs). This paper begins by establishing a PHEV model based on physical and data-driven models, focusing on the high-fidelity training environment. It then proposes a real-vehicle application-oriented control framework, combining horizon-extended reinforcement learning (RL)-based energy management with the equivalent consumption minimization strategy (ECMS) to enhance practical applicability, and improves the flawed method of equivalent factor evaluation based on instantaneous driving cycle and powertrain states found in existing research. Finally, comprehensive simulation and hardware-in-the-loop validation are carried out which demonstrates the advantages of the proposed control framework in fuel economy over adaptive-ECMS and rule-based strategies. Compared to conventional RL architectures that directly control powertrain components, the proposed control method not only achieves similar optimality but also significantly enhances the disturbance resistance of the energy management system, providing an effective control framework for RL-based energy management strategies aimed at real-vehicle applications by OEMs.

The form factors of $B_c\rightarrow\eta_c$ and $B_c\rightarrow J/\psi$ are analyzed in the framework of three-point QCD sum rules. In these analyses, the contributions of the vacuum condensate terms $\langle g_{s}^{2}GG\rangle$ and $\langle g_{s}^{3}GGGf\rangle$ are considered. In addition, the decay widths and branching ratios of several decay channels are obtained by using the calculated form factors. These decay processes include the nonleptonic decays of $B_c^- \to \eta_c \pi^-$, $\eta_c K^-$, $\eta_c \rho^-$, $\eta_c K^{*-}$, $B_c^- \to J/\psi \pi^-$, $J/\psi K^-$, $J/\psi \rho^-$, $J/\psi K^{*-}$, and the semileptonic decays of $B_c^- \to \eta_c \mathcal{l} \bar{\nu}$, $B_c^- \to J/\psi \mathcal{l} \bar{\nu}$. These results about the form factors and decay properties of $B_c$ meson provide useful information for us to study the heavy-quark dynamics and find new physics(NP) beyond Standard Model(SM).

Fine-Tuning Diffusion Models enable a wide range of personalized generation and editing applications on diverse visual modalities. While Low-Rank Adaptation (LoRA) accelerates the fine-tuning process, it still requires multiple reference images and time-consuming training, which constrains its scalability for large-scale and real-time applications. In this paper, we propose \textit{View Iterative Self-Attention Control (VisCtrl)} to tackle this challenge. Specifically, VisCtrl is a training-free method that injects the appearance and structure of a user-specified subject into another subject in the target image, unlike previous approaches that require fine-tuning the model. Initially, we obtain the initial noise for both the reference and target images through DDIM inversion. Then, during the denoising phase, features from the reference image are injected into the target image via the self-attention mechanism. Notably, by iteratively performing this feature injection process, we ensure that the reference image features are gradually integrated into the target image. This approach results in consistent and harmonious editing with only one reference image in a few denoising steps. Moreover, benefiting from our plug-and-play architecture design and the proposed Feature Gradual Sampling strategy for multi-view editing, our method can be easily extended to edit in complex visual domains. Extensive experiments show the efficacy of VisCtrl across a spectrum of tasks, including personalized editing of images, videos, and 3D scenes., Comment: Under review

This work establishes the time-dependent relativistic Hartree-Fock (TD-RHF) model with spherical symmetry for the first time. The time-dependent integro-differential Dirac equations are solved by expanding Dirac spinors on the spherical Dirac Woods-Saxon (DWS) basis. The numerical verification demonstrates the high conservation qualities for both the total binding energy and the particle number, as well as the time-reversal invariance of the system, which ensures the precision and reliability of the newly developed TD-RHF model. Subsequently, the isoscalar giant monopole resonance (ISGMR) mode of $^{208}$Pb is investigated using the RHF Lagrangian PKO1. The constrained energy of the ISGMR calculated by PKO1 is found to be in close agreement with the experimental data, and the strength function is similar to the results given by the relativistic Hartree-Fock plus random phase approximation. Based on the advantage of the TD-RHF model in avoiding complicated calculations of the residual interactions, the ISGMR mode of $^{208}$Pb is calculated by twelve relativistic effective Lagrangians. The results indicate that the value of the incompressibility of nuclear matter $K_\infty$ constrained by relativistic effective Lagrangians is in the range of $237\sim246$ MeV, which is lower than the previous investigations based on the relativistic models., Comment: 9 pages, 5 figures

Materials that exhibit strongly coupled magnetic order and electronic properties are crucial for both fundamental research and technological applications. However, finding a material that not only shows remarkable magnetoresistive responses but also has an easily tunable ground state remains a challenge. Here, we report successful manipulation of the magnetic and transport properties of EuCd$_2$P$_2$, which is transformed from an A-type antiferromagnet ($T_\mathrm{N}$ = 11 K) exhibiting colossal magnetoresistance into a ferromagnet ($T_\mathrm{C}$ = 47 K) with metallic behavior. The dramatic alteration results from a low hole concentration of $10^{19}$ cm$^{-3}$ induced by changing the growth conditions. Electronic structure and total energy calculations confirm the tunability of magnetism with a small carrier concentration for EuCd$_2$P$_2$. It is feasible to switch between the magnetic states by using field-effect to control the carrier density, thereby changing the magneto-electronic response. The controllable magnetism and electrical transport of EuCd$_2$P$_2$ make it a potential candidate for spintronics.

EuZn$_2$P$_2$ was reported to be an insulating antiferromagnet with $T_\mathrm{N}$ of 23.5 K. In this study, single crystals of EuZn$_2$P$_2$ exhibiting metallic behavior and a ferromagnetic order of 72 K ($T_\mathrm{C}$) are successfully synthesized via a salt flux method. The presence of hole carriers induced by the Eu vacancies in the lattice is found to be crucial for the drastic changes in magnetism and electrical transport. The carriers mediate the interlayer ferromagnetic interaction, and the coupling strength is directly related to $T_\mathrm{C}$, as evidenced by the linear dependence of $T_\mathrm{C}$ and the fitted Curie-Weiss temperatures on the Eu-layer distances for ferromagnetic Eu$M_2X_2$ ($M$ = Zn, Cd; $X$ = P, As). The ferromagnetic EuZn$_2$P$_2$ shows conspicuous negative magnetoresistance (MR) near $T_\mathrm{C}$, owing to strong magnetic scattering. The MR behavior is consistent with the Majumdar-Littlewood model, indicating that the MR can be enhanced by decreasing the carrier density. Our findings suggest that Eu$M_2X_2$ has highly tunable magnetism and charge transport, making it a promising material family for potential applications in spintronics.

Despite progress, deep neural networks still suffer performance declines under distribution shifts between training and test domains, leading to a substantial decrease in Quality of Experience (QoE) for multimedia applications. Existing test-time adaptation (TTA) methods are challenged by dynamic, multiple test distributions within batches. This work provides a new perspective on analyzing batch normalization techniques through class-related and class-irrelevant features, our observations reveal combining source and test batch normalization statistics robustly characterizes target distributions. However, test statistics must have high similarity. We thus propose Discover Your Neighbours (DYN), the first backward-free approach specialized for dynamic TTA. The core innovation is identifying similar samples via instance normalization statistics and clustering into groups which provides consistent class-irrelevant representations. Specifically, Our DYN consists of layer-wise instance statistics clustering (LISC) and cluster-aware batch normalization (CABN). In LISC, we perform layer-wise clustering of approximate feature samples at each BN layer by calculating the cosine similarity of instance normalization statistics across the batch. CABN then aggregates SBN and TCN statistics to collaboratively characterize the target distribution, enabling more robust representations. Experimental results validate DYN's robustness and effectiveness, demonstrating maintained performance under dynamic data stream patterns., Comment: 10 pages

Nowadays, large language models (LLMs) are published as a service and can be accessed by various applications via APIs, also known as language-model-as-a-service (LMaaS). Without knowing the generation length of requests, existing serving systems serve requests in a first-come, first-served (FCFS) manner with a fixed batch size, which leads to two problems that affect batch serving efficiency. First, the generation lengths of requests in a batch vary, and requests with short generation lengths must wait for requests with long generation lengths to finish during the batch serving procedure. Second, requests with longer generation lengths consume more memory during serving. Without knowing the generation lengths of batched requests, the batch size is always set small to avoid the out-of-memory (OOM) error, thus preventing the GPU from being fully utilized. In this paper, we find that a significant number of popular applications in the LMaaS scenario have a positive correlation between the generation length and the length of raw user input. Based on this observation, we propose Magnus, which can accurately predict the request generation length with the user input length, application-level, and user-level semantic features. Accordingly, Magnus can achieve high request throughput by batching requests of similar generation lengths together with adaptive batch sizes. Besides, Magnus can also schedule batches with the highest response ratio next (HRRN) policy to reduce request response time. Experiments conducted on our testbed show that Magnus improves request throughput by up to 234\% and reduces response time by up to 89.7\% compared to baselines., Comment: 12 pages, 14 figures