arXiv subjects

Qiang Luo

Publications and source records attributed to Qiang Luo.

At least 55 records · Page 3Linked to original sources

arXiv2025

Emergent supercounterfluid and quantum phase diagram of two-component interacting bosons in one-dimensional optical lattice

Motivated by a recent experiment that realizes nearest-neighbor dipolar couplings in an optical lattice [C. Lagoin, $\textit{et al.}$, Nature $\textbf{609}$, 485 (2022)], we study a one-dimensional version of the two-component extended Bose-Hubbard model via the density-matrix renormalization group method. By using the nearest-neighbor and on-site interaction parameters from the experiment, we start by mapping the quantum phase diagram in the hopping parameters $t_{A}\mbox{-}t_{B}$ plane with boson densities $ρ_{A}=ρ_{B}=1/2$. In addition to the density wave phase reported in the experiment, we find several regimes of superfluidity when one or two hopping parameters are large enough, and interestingly there is a supercounterfluid phase at moderate and comparable hopping parameters. The universality classes of these phase transitions are analyzed from the correlation functions, excitation gaps, and entanglement entropy. In particular, a Berezinskii-Kosterlitz-Thouless type is recognized several gapped-to-gapless transitions. In addition, we also study the quantum phase transitions when varying $ρ_{B}$ from 0 to 1 while keeping $ρ_A = 1/2$. We identify a supersolid phase in a wide range of $1/2<ρ_B<1$. Our work paves the way for realizing exotic many-body phases in cold atom experiments upon proper tuning of experimental parameters.

cond-mat.quant-gas↗

arXiv2025

Differential Privacy Personalized Federated Learning Based on Dynamically Sparsified Client Updates

Personalized federated learning is extensively utilized in scenarios characterized by data heterogeneity, facilitating more efficient and automated local training on data-owning terminals. This includes the automated selection of high-performance model parameters for upload, thereby enhancing the overall training process. However, it entails significant risks of privacy leakage. Existing studies have attempted to mitigate these risks by utilizing differential privacy. Nevertheless, these studies present two major limitations: (1) The integration of differential privacy into personalized federated learning lacks sufficient personalization, leading to the introduction of excessive noise into the model. (2) It fails to adequately control the spatial scope of model update information, resulting in a suboptimal balance between data privacy and model effectiveness in differential privacy federated learning. In this paper, we propose a differentially private personalized federated learning approach that employs dynamically sparsified client updates through reparameterization and adaptive norm(DP-pFedDSU). Reparameterization training effectively selects personalized client update information, thereby reducing the quantity of updates. This approach minimizes the introduction of noise to the greatest extent possible. Additionally, dynamic adaptive norm refers to controlling the norm space of model updates during the training process, mitigating the negative impact of clipping on the update information. These strategies substantially enhance the effective integration of differential privacy and personalized federated learning. Experimental results on EMNIST, CIFAR-10, and CIFAR-100 demonstrate that our proposed scheme achieves superior performance and is well-suited for more complex personalized federated learning scenarios.

cs.LG↗

arXiv2025

OneRec: Unifying Retrieve and Rank with Generative Recommender and Iterative Preference Alignment

Recently, generative retrieval-based recommendation systems have emerged as a promising paradigm. However, most modern recommender systems adopt a retrieve-and-rank strategy, where the generative model functions only as a selector during the retrieval stage. In this paper, we propose OneRec, which replaces the cascaded learning framework with a unified generative model. To the best of our knowledge, this is the first end-to-end generative model that significantly surpasses current complex and well-designed recommender systems in real-world scenarios. Specifically, OneRec includes: 1) an encoder-decoder structure, which encodes the user's historical behavior sequences and gradually decodes the videos that the user may be interested in. We adopt sparse Mixture-of-Experts (MoE) to scale model capacity without proportionally increasing computational FLOPs. 2) a session-wise generation approach. In contrast to traditional next-item prediction, we propose a session-wise generation, which is more elegant and contextually coherent than point-by-point generation that relies on hand-crafted rules to properly combine the generated results. 3) an Iterative Preference Alignment module combined with Direct Preference Optimization (DPO) to enhance the quality of the generated results. Unlike DPO in NLP, a recommendation system typically has only one opportunity to display results for each user's browsing request, making it impossible to obtain positive and negative samples simultaneously. To address this limitation, We design a reward model to simulate user generation and customize the sampling strategy. Extensive experiments have demonstrated that a limited number of DPO samples can align user interest preferences and significantly improve the quality of generated results. We deployed OneRec in the main scene of Kuaishou, achieving a 1.6\% increase in watch-time, which is a substantial improvement.

cs.IR↗

arXiv2025

FARM: Frequency-Aware Model for Cross-Domain Live-Streaming Recommendation

Live-streaming services have attracted widespread popularity due to their real-time interactivity and entertainment value. Users can engage with live-streaming authors by participating in live chats, posting likes, or sending virtual gifts to convey their preferences and support. However, the live-streaming services faces serious data-sparsity problem, which can be attributed to the following two points: (1) User's valuable behaviors are usually sparse, e.g., like, comment and gift, which are easily overlooked by the model, making it difficult to describe user's personalized preference. (2) The main exposure content on our platform is short-video, which is 9 times higher than the exposed live-streaming, leading to the inability of live-streaming content to fully model user preference. To this end, we propose a Frequency-Aware Model for Cross-Domain Live-Streaming Recommendation, termed as FARM. Specifically, we first present the intra-domain frequency aware module to enable our model to perceive user's sparse yet valuable behaviors, i.e., high-frequency information, supported by the Discrete Fourier Transform (DFT). To transfer user preference across the short-video and live-streaming domains, we propose a novel preference align before fuse strategy, which consists of two parts: the cross-domain preference align module to align user preference in both domains with contrastive learning, and the cross-domain preference fuse module to further fuse user preference in both domains using a serious of tailor-designed attention mechanisms. Extensive offline experiments and online A/B testing on Kuaishou live-streaming services demonstrate the effectiveness and superiority of FARM. Our FARM has been deployed in online live-streaming services and currently serves hundreds of millions of users on Kuaishou.

cs.IR↗

arXiv2025

Soliton microcombs in X-cut LiNbO3 microresonators

Chip-scale integration of optical frequency combs, particularly soliton microcombs, enables miniaturized instrumentation for timekeeping, ranging, and spectroscopy. Although soliton microcombs have been demonstrated on various material platforms, realizing complete comb functionality on photonic chips requires the co-integration of high-speed modulators and efficient frequency doublers, features that are available in a monolithic form on X-cut thin-film lithium niobate (TFLN). However, the pronounced Raman nonlinearity associated with extraordinary light in this platform has so far precluded soliton microcomb generation. Here, we report the generation of transverse-electric-polarized soliton microcombs with a 25 GHz repetition rate in high-Q microresonators on X-cut TFLN chips. By precisely orienting the racetrack microresonator relative to the optical axis, we mitigate Raman nonlinearity and enable soliton formation under continuous-wave laser pumping. Moreover, the soliton microcomb spectra are extended to 350 nm with pulsed laser pumping. This work expands the capabilities of TFLN photonics and paves the way for the monolithic integration of fast-tunable, self-referenced microcombs.

physics.optics↗

arXiv2025

Successive topological phase transitions in two distinct spin-flop phases on the honeycomb lattice

The Kitaev magnets with bond-dependent interactions have garnered considerable attention in recent years for their ability to harbor exotic phases and nontrivial excitations. The topological magnons, which are indicated by nonzero Chern number that can enhance the thermal Hall conductivity, are proposed to partially explain thermal Hall measurements in real materials. Hitherto, topological magnons have been extensively explored when the magnetic field is normal to the honeycomb plane, but their topological characteristics are less studied in the presence of in-plane magnetic field. Here, we study two distinct in-plane field induced spin-flop phases in the $Γ$-$Γ'$ model, both of which are off-diagonal couplings that have intimate relation to the Kitaev interaction. The two spin-flop phases are distinguished by their out-of-plane spin components which can be either antiparallel or parallel, thus dubbing antiferromagnetic (AFM) or ferromagnetic (FM) spin-flop phases, respectively. We map out topological phase diagrams for both phases, revealing a rich pattern of the Chern number over exchange parameters and magnetic field. We analytically calculate the boundaries of topological phase transitions when the magnetic field is along the $a$ and $b$ directions. We find that the thermal Hall conductivity and its derivative display contrasting behaviors when crossing different topological phase transitions. The striking difference of the two phases lies in that when the magnetic field is along the $b$ direction, topological magnons are totally absent in the AFM spin-flop phase, while they can survive in the FM analogue in certain parameter regions.

cond-mat.str-el↗

arXiv2024

Interplay of Kitaev Interaction and Off-diagonal Exchanges: Exotic Phases and Quantum Phase Diagrams

Aligning with the everlasting search for quantum spin liquids (QSLs), identifying the QSL in Kitaev magnets has garnered great research interest during the past decade and remains nevertheless an enormous challenge. One of the major difficulties lies in that Kitaev QSL is typically fragile against competing interactions like off-diagonal exchanges, which are ubiquitous in real materials due to spin-orbit coupling and crystal-field effect. This, in turn, gives rise to many intriguing field-induced novel phases and thermal Hall effect. In this review, we will focus on the interplay of Kitaev interaction and off-diagonal $Γ$ and $Γ'$ exchanges from a numerical perspective. This review discusses some representative exotic phases such as $Γ$ spin liquid, nematic ferromagnet, spin-flop phase, and distinct chiral-spin states with spontaneously time-reversal symmetry breaking. It also presents quantum phase diagrams of anisotropic Kitaev-$Γ$ chains that exhibit kaleidoscopes of both ordered and disordered phases.

cond-mat.str-el↗

arXiv2024

Commensurate to Incommensurate Transition of Three Dimensional Charge Density Waves

Charge density wave (CDW) is a widely concerned emergent phenomenon in condensed matter physics. To establish a systematic understanding of CDW, we develop a diagrammatic self-consistent-field approach for cubic Holstein model employing fluctuation exchange approximation, and explore the emergence and transition of three-dimensional CDWs. Commensurate CDW (c-CDW) locked at $(π,π,π)$ is favored near half-filling, and the transition temperature is predicted around half of the nearest-neighbor hopping. Large hole doping leads to a suppression of CDW transition temperature and the emergence of incommensurate CDW (i-CDW), which is evidenced by a drifting of the ordering vector away from $(π,π,π)$ towards $(π,π,0)$. Phonon frequency significantly impacts the transition temperature and the phase boundary between c-CDW and i-CDW, and the optimal frequency for enlarging the CDW regime is also predicted near half of the nearest-neighbor hopping. These new theoretical results provide a systematic understanding of CDW and a fresh perspective on emergent phenomena dominated by electron-phonon interaction.

cond-mat.str-el↗

arXiv2024

KuaiFormer: Transformer-Based Retrieval at Kuaishou

In large-scale content recommendation systems, retrieval serves as the initial stage in the pipeline, responsible for selecting thousands of candidate items from billions of options to pass on to ranking modules. Traditionally, the dominant retrieval method has been Embedding-Based Retrieval (EBR) using a Deep Neural Network (DNN) dual-tower structure. However, applying transformer in retrieval tasks has been the focus of recent research, though real-world industrial deployment still presents significant challenges. In this paper, we introduce KuaiFormer, a novel transformer-based retrieval framework deployed in a large-scale content recommendation system. KuaiFormer fundamentally redefines the retrieval process by shifting from conventional score estimation tasks (such as click-through rate estimate) to a transformer-driven Next Action Prediction paradigm. This shift enables more effective real-time interest acquisition and multi-interest extraction, significantly enhancing retrieval performance. KuaiFormer has been successfully integrated into Kuaishou App's short-video recommendation system since May 2024, serving over 400 million daily active users and resulting in a marked increase in average daily usage time of Kuaishou users. We provide insights into both the technical and business aspects of deploying transformer in large-scale recommendation systems, addressing practical challenges encountered during industrial implementation. Our findings offer valuable guidance for engineers and researchers aiming to leverage transformer models to optimize large-scale content recommendation systems.

cs.IR↗

arXiv2024

Network Structure Governs Drosophila Brain Functionality

How intelligence emerges from living beings has been a fundamental question in neuroscience. However, it remains largely unanswered due to the complex neuronal dynamics and intricate connections between neurons in real neural systems. To address this challenge, we leveraged the largest available adult Drosophila connectome data set, and constructed a comprehensive computational framework based on simplified neuronal activation mechanisms to simulate the observed activation behavior within the connectome. The results revealed that even with rudimentary neuronal activation mechanisms, models grounded in real neural network structures can generate activation patterns strikingly similar to those observed in the actual brain. A significant discovery was the consistency of activation patterns across various neuronal dynamic models. This consistency, achieved with the same network structure, underscores the pivotal role of network topology in neural information processing. These results challenge the prevailing view that solely relies on neuron count or complex individual neuron dynamics. Further analysis demonstrated a near-complete separation of the visual and olfactory systems at the network level. Moreover, we found that the network distance, rather than spatial distance, is the primary determinant of activation patterns. Additionally, our experiments revealed that a reconnect rate of at least 0.1% was sufficient to disrupt the previously observed activation patterns. We also observed synergistic effects between the brain hemispheres: Even with unilateral input stimuli, visual-related neurons in both hemispheres were activated, highlighting the importance of interhemispheric communication. These findings emphasize the crucial role of network structure in neural activation and offer novel insights into the fundamental principles governing brain functionality.

q-bio.NC↗

arXiv2024

Chiral spin state and nematic ferromagnet in the spin-1 Kitaev-$Γ$ model

The higher-spin Kitaev magnets, in which the Kitaev interaction and off-diagonal exchange couplings are overwhelmingly large, have emerged as a fertile avenue to explore exotic phases and unusual excitations. In this work, we study the quantum phase diagram of the spin-1 Kitaev-$Γ$ model on the honeycomb lattice using density-matrix renormalization group. It harbours six distinct phases and the intriguing findings are three magnetically ordered phases in which both time-reversal symmetry and lattice symmetry albeit of different sort are broken spontaneously. The chiral spin state originates from the order-by-disorder effect and exhibits an almost saturated scalar spin chirality at the quantum level. Depending on the relative strength of the two interactions, it also features columnar-like or plaquette-like dimer pattern as a consequence of the translational symmetry breaking. In parallel, the nematic ferromagnets are situated at ferromagnetic Kitaev side and possess small but finite ferromagnetic ordering. The lattice-rotational symmetry breaking enforces nonequivalent bond energy along one of the three bonds. Although the intrinsic difference between the two nematic ferromagnets remains elusive, the discontinuities in the von Neumann entropy, hexagonal plaquette operator, and Wilson loop operator convincingly suggest that they are separated via a first-order phase transition.

cond-mat.str-el↗

arXiv2024

Topological phase transitions and thermal Hall effect in a noncollinear spin texture

The noncollinear spin textures provide promising avenues to stabilize exotic magnetic phases and excitations. They have attracted vast attention owning to their nontrivial band topology in the past decades. Distinct from the conventional route of involving the Dzyaloshinskii-Moriya interaction in a honeycomb magnet, the interplay of bond-dependent Kitaev and $Γ$ interactions, originating from the spin-orbit coupling and octahedra crystal field in real materials, has demonstrated to be another source to generate noncollinear spin textures with multiple spins in a magnetic unit cell. Notably, earlier works have revealed a triple-meron crystal (TmX) consisting of eighteen spins in the frustrated Kitaev-$Γ$ model. Aligning with previous efforts, here we attempt to identify that the TmX hosts several peculiar features with the help of the linear spin-wave theory. To begin with, the symmetric anisotropic exchanges are beneficial for the existence of nonreciprocal magnons, which are stabilized by an external magnetic field. Further, within the regime of TmX, successive topological phase transitions occur, accompanied by the changes of Chern number in value and thermal Hall conductivity in sign. In addition, topological nature of magnons is also verified by the onset of chiral edge modes in a nanoribbon geometry. Our findings pave the way to study topological phenomena of noncollinear spin textures in potential Kitaev materials.

cond-mat.str-el↗

arXiv2024

DemiNet: Dependency-Aware Multi-Interest Network with Self-Supervised Graph Learning for Click-Through Rate Prediction

In this paper, we propose a novel model named DemiNet (short for DEpendency-Aware Multi-Interest Network) to address the above two issues. To be specific, we first consider various dependency types between item nodes and perform dependency-aware heterogeneous attention for denoising and obtaining accurate sequence item representations. Secondly, for multiple interests extraction, multi-head attention is conducted on top of the graph embedding. To filter out noisy inter-item correlations and enhance the robustness of extracted interests, self-supervised interest learning is introduced to the above two steps. Thirdly, to aggregate the multiple interests, interest experts corresponding to different interest routes give rating scores respectively, while a specialized network assigns the confidence of each score. Experimental results on three real-world datasets demonstrate that the proposed DemiNet significantly improves the overall recommendation performance over several state-of-the-art baselines. Further studies verify the efficacy and interpretability benefits brought by the fine-grained user interest modeling.

cs.IR↗

arXiv2024

Beyond Kitaev physics in strong spin-orbit coupled magnets

We review the recent advances and current challenges in the field of strong spin-orbit coupled Kitaev materials, with a particular emphasis on the physics beyond the exactly-solvable Kitaev spin liquid point. To that end, we give a comprehensive overview of the most relevant exchange interactions in $d^5$ and $d^7$ iridates and similar compounds, an exposition of their microscopic origin, and a systematic attempt to map out the most interesting correlated regimes of the multi-dimensional parameter space, guided by powerful symmetry and duality transformations as well as by insights from wide-ranging analytical and numerical studies. We also survey recent exciting results on quasi-1D models and discuss their relevance to higher-dimensional models. Finally, we highlight some of the key questions in the field as well as future directions.

cond-mat.str-el↗

arXiv2023

High-efficiency edge couplers enabled by vertically tapering on lithium-niobate photonic chips

In the past decade, photonic integrated circuits (PICs) based on thin-film lithium niobate (TFLN) have advanced in various fields, including optical communication, nonlinear photonics, and quantum optics. A critical component is an efficient edge coupler connecting PICs to light sources or detectors. Here, we propose an innovative edge coupler design with a wedge-shaped TFLN waveguide and a silicon oxynitride (SiON) cladding. Experimental results show that the coupling loss between the TFLN PIC and a 3-μm mode field diameter (MFD) lensed fiber is low at 1.52 dB/facet, with the potential for improvement to 0.43 dB/facet theoretically. The coupling loss between the edge coupler and a UHNA7 fiber with an MFD of 3.2 μm is reduced to 0.92 dB/facet. This design maintains robust fabrication and alignment tolerance. Importantly, the minimum linewidth of the TFLN waveguide of the coupler (600 nm) can be easily achieved using foundry-available i-line stepper lithography. This work benefits the development of TFLN integrated platforms, such as on-chip electro-optic modulators, frequency comb generation, and quantum sensors.

physics.optics↗

arXiv2023

Evidence of Kitaev interaction in the monolayer 1T-CrTe$_2$

The two-dimensional 1T-CrTe$_2$ has been an attractive room-temperature van der Waals magnet which has a potential application in spintronic devices. Although it was recognized as a ferromagnetism in the past, the monolayer 1T-CrTe$_2$ was recently found to exhibit zigzag antiferromagnetism with the easy axis oriented at $70^\circ$ to the perpendicular direction of the plane. Therefore, the origin of the intricate anisotropic magnetic behavior therein is well worthy of thorough exploration. Here, by applying density functional theory with spin spiral method, we demonstrate that the Kitaev interaction, together with the single-ion anisotropy and other off-diagonal exchanges, is amenable to explain the magnetic orientation in the metallic 1T-CrTe$_2$. Moreover, the Ruderman-Kittle-Kasuya-Yosida interaction can also be extracted from the dispersion calculations, which explains the metallic behavior of 1T-CrTe$_2$. Our results demonstrate that 1T-CrTe$_2$ is potentially a rare metallic Kitaev material.

cond-mat.str-el↗

arXiv2023

Spontaneous dimerization, spin-nematic order, and deconfined quantum critical point in a spin-1 Kitaev chain with tunable single-ion anisotropy

The Kitaev-type spin chains have been demonstrated to be fertile playgrounds in which exotic phases and unconventional phase transitions are ready to appear. In this work, we use the density-matrix renormalization group method to study the quantum phase diagram of a spin-1 Kitaev chain with a tunable negative single-ion anisotropy (SIA). When the strength of the SIA is small, the ground state is revealed to be a spin-nematic phase which escapes conventional magnetic order but is characterized by a finite spin-nematic correlation because of the breaking spin-rotational symmetry. As the SIA increases, the spin-nematic phase is taken over by either a dimerized phase or an antiferromagnetic phase through an Ising-type phase transition, depending on the direction of the easy axis. For large enough SIA, the dimerized phase and the antiferromagnetic phase undergo a ``Landau-forbidden" continuous phase transition, suggesting new platform of deconfined quantum critical point in spin-1 Kitaev chain.

cond-mat.str-el↗

arXiv2022

Interplay of magnetic field and trigonal distortion in the honeycomb $Γ$ model: Occurrence of a spin-flop phase

In candidate Kitaev materials, the off-diagonal $Γ$ and $Γ'$ interactions are identified to come from the spin-orbit coupling and trigonal distortion, respectively. They have generated intense research efforts because of their intimate relation to the field-induced magnetically disordered state reported in $α$-RuCl$_3$. Theoretically, while a plethora of field-induced phases has been proposed in the honeycomb lattice, a stable intermediate phase that can survive in a wide parameter region regardless of the underlying phases is still lacking. Here we focus on the interplay of an out-of-plane magnetic field and a symmetry-allowed $Γ'$ term due to trigonal distortion in the dominant antiferromagnetic $Γ$ region. By using multifaceted approaches ranging from classical Monte Carlo and semiclassical spin-wave theory to density-matrix renormalization group, we identify an intriguing spin-flop phase in the presence of magnetic field and antiferromagnetic $Γ'$ interaction, before it eventually enters into a fully polarized state. As the $Γ'$ interaction approaches the size of $Γ$ one, the $Γ$-$Γ'$ model maps to the easy-axis XXZ antiferromagnet, where the spin-flop phase can be understood as a superfluid phase in the extended Bose-Hubbard model. Our finding thus demonstrates an exciting path from the honeycomb $Γ$ model towards a $U(1)$-symmetric XXZ antiferromagnet in a magnetic field.

cond-mat.str-el↗