arXiv subjects

Hongbin Zhang

Publications and source records attributed to Hongbin Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2025

Shift Current Anomalous Photovoltaics in a Double Perovskite Ferroelectric

Ferroelectric anomalous photovoltaic (APV) effect, as a fascinating physical conceptual phenomenon, holds significant potentials for new optoelectronic device applications. However, due to the knowledge lacking on the origin and underlying mechanism of ferroelectric APV effect, substantial challenges still remain in exploring new APV-active candidate materials. The emerging shift current model, involving the transfer of photogenerated charges through the displacement of wave functions, has attracted considerable attention for its unique insights into the bulk photovoltaic effect. Here, we present strong APV properties in a high-temperature double perovskite ferroelectric (cyclohexylmethylammonium)$_2$CsAgBiBr$_7$, showing an extremely large above-bandgap photovoltage up to about 40 V. This figure-of-merit is far beyond its bandgap of about 2.3 eV and comparable to the state-of-art molecular ferroelectrics. Strikingly, the shift current model reveals an intrinsic correlation with Cs$^{+}$ cation displacement and provides, for the first time, an explicit explanation for the structural origin of ferroelectric APV activities. Besides, its steady-state APV photocurrent exhibits the unique light-polarization dependence, which endows remarkable polarization-sensitivity with the highest polarization ratios of about 41 among the known 2D single-phase materials. As the unprecedented exploration of ferroelectric APV characteristics illuminated by the shift current mechanism, this finding paves a pathway to assemble new optoelectronic smart devices.

cond-mat.mtrl-sci↗

arXiv2025

Optimizing $α''$-Fe$_{16}$N$_2$ as permanent magnet via alloying

Based on systematic first-principles calculations, we investigate the effects of 27 alloying elements on the intrinsic magnetic properties of Fe$_{16}$N$_2$, in order to further optimize its properties for permanent magnet applications. Analysis on the thermodynamic stabilities based on formation energy and distance to the convex hull reveals that 20 elements can be substituted into Fe$_{16}$N$_2$, where there is no strong site-preference upon doping. It is observed that all alloying elements can essentially reduce the saturation magnetization, whereas the magnetic anisotropy can be significantly modified. In terms of the Boltzmann-average intrinsic properties, we identify 8 elements as interesting candidates, with Co, Mo, and W as the most promising cases for further experimental validations.

cond-mat.mtrl-sci↗

arXiv2025

TD-Pipe: Temporally-Disaggregated Pipeline Parallelism Architecture for High-Throughput LLM Inference

As the model size continuously increases, pipeline parallelism shows great promise in throughput-oriented LLM inference due to its low demand on communications. However, imbalanced pipeline workloads and complex data dependencies in the prefill and decode phases result in massive pipeline bubbles and further severe performance reduction. To better exploit the pipeline parallelism for high-throughput LLM inference, we propose TD-Pipe, with the key idea lies in the temporally-disaggregated pipeline parallelism architecture. Specifically, this architecture disaggregates the prefill and decode phases in the temporal dimension, so as to eliminate pipeline bubbles caused by the phase switching. TD-Pipe identifies potential issues of exploiting the novel architecture and provides solutions. First, a hierarchy-controller structure is used to better coordinate devices in pipeline parallelism by decoupling the scheduling from execution. Second, the AI-based greedy prefill approach aggressively performs more prefills by predicting the output length and simulating the memory usage. Third, the inter-batch work stealing approach dynamically balances decode phase workloads between different batches to reduce bubbles. Forth, the spatial-temporal intensity comparison approach determines the optimal switch from decode to prefill by comparing the performance drop from reduced computational intensity with that from phase switching bubbles. Extensive experiments show that TD-Pipe effectively increases the throughput of LLM inference by up to 1.91x over the existing tensor parallel approach and 2.73x over the existing pipeline parallel approach on GPU nodes with only PCIe interconnection.

cs.DC↗

arXiv2025

EcoServe: Enabling Cost-effective LLM Serving with Proactive Intra- and Inter-Instance Orchestration

Existing LLM serving strategies can be categorized based on whether prefill and decode phases are disaggregated: non-disaggregated (NoDG) or fully disaggregated (FuDG). However, the NoDG strategy leads to strong prefill-decode interference and the FuDG strategy highly relies on high-performance interconnects, making them less cost-effective. We introduce EcoServe, a system that enables cost-effective LLM serving on clusters with commodity interconnects. EcoServe is built on the partially disaggregated (PaDG) strategy, applying temporal disaggregation and rolling activation for proactive intra- and inter-instance scheduling. It first disaggregates the prefill and decode phases along the time dimension within a single instance to mitigate inter-phase interference and enhance throughput. Next, it coordinates multiple instances and cyclically activates them to ensure the continuous availability of prefill processing, thereby improving latency. Thus, EcoServe's basic serving unit is the macro instance, within which multiple instances collaborate. It further integrates an adaptive scheduling algorithm to route requests in a macro instance and a mitosis scaling approach to enable fine-grained capacity scaling. Beyond delivering high goodput, EcoServe excels in load balancing, hardware cost, parallelism compatibility, and even engineering simplicity compared to existing solutions. When serving 30B- and 70B-scale models on a production-level cluster with 32 NVIDIA L20 GPUs using commodity Ethernet, EcoServe averagely improves goodput by 82.49%, 86.17%, 122.76%, and 126.96% over four representative NoDG and FuDG systems.

cs.DC↗

arXiv2025

Quantum correlation and origin of Hawking radiation for mass-superposed BTZ black holes

We investigate the mass superposition of BTZ black holes using the velocity correlation function. Interestingly, the quantum signatures of BTZ black hole mass superposition can be revealed by the velocity correlation peaks. Additionally, different mass superpositions for the same mass ratio can be distinguished, a phenomenon not previously documented. We also find that the correlation function method can pinpoint the location where Hawking radiations are generated, even for masssuperposed BTZ black holes. This supports the concept of a quantum atmosphere as the origin locus of Hawking radiation, even in quantum-superposed spacetime.

hep-th↗

arXiv2025

High-throughput computation and machine learning modeling of magnetic moments and Mössbauer parameters for Fe-based intermetallics

Based on high-throughput density functional theory calculations, we evaluate the local magnetic moments and Mössbauer properties for Fe-based intermetallic compounds and employ machine learning to map the local crystalline environments to such properties. It is observed that magnetic moments and Mössbauer parameters provide complementary insights into the local crystalline environment, where the statistical features cannot be captured using phenomenological models. Furthermore, we find that the scarcity of existing data in Materials Project (MP) poses a significant challenge in developing predictive machine learning models, whereas SOAP-based descriptors can be applied for reliable modeling of the enriched datasets with extra structural prototypes. This work advances the mapping of local crystalline structures to magnetic and spectroscopic properties, bridging the gap between empirical observations and theoretical models.

cond-mat.mtrl-sci↗

arXiv2025

LinguaLIFT: An Effective Two-stage Instruction Tuning Framework for Low-Resource Language Reasoning

Large language models (LLMs) have exhibited impressive multilingual reasoning capabilities, driven by extensive multilingual pre-training corpora and instruction fine-tuning data. However, a performance gap exists between high- and low-resource language reasoning tasks due to the language imbalance in the pre-training corpus, which is exacerbated by evaluation bias in existing reasoning benchmarks lacking low-resource language coverage. To alleviate this issue, we propose LinguaLIFT, a two-stage instruction tuning framework for advancing low-resource language reasoning. LinguaLIFT employs a language alignment layer to capture multilingual alignment in a code-switched tuning way without requiring multilingual instruction or parallel data, thereby transferring the cross-lingual reasoning capabilities to low-resource languages through English-only instruction tuning data. To comprehensively evaluate the multilingual reasoning capabilities, we introduce the Multilingual Math World Problem (MMWP) benchmark, which spans 21 low-resource, 17 medium-resource, and 10 high-resource languages. Experimental results show that LinguaLIFT outperforms several competitive baselines across MMWP and four widely used benchmarks.

cs.CL↗

arXiv2025

Voltage-driven 90 switching of bulk perpendicular magnetic anisotropy in ferrimagnets

Unravelling the mechanism behind bulk perpendicular magnetic anisotropy (PMA) in amorphous rare earth-transition metal films has proven challenging. This is largely due to the inherent complexity of amorphous structure and the entangled potential origins arising from microstructure and atomic structure factors. Here, we present an approach wherein the magneto-electric effect is harnessed to induce 90° switching of bulk PMA in Tb-Co films to in-plane directions by applying voltages of only -1.2 V. This manipulation is achieved by voltage-driven insertion of hydrogen atoms into interstitial sites between Tb and Co atoms, which serves as a perturbation to local atomic structure. Using angle-dependent X-ray magnetic circular dichroism, we find that the anisotropy switching originates from the distortion of crystal field around Tb which reorients the alignment of Tb orbital moments. Initially aligned along Tb-Co bonding directions, the easy magnetization axis undergoes reorientation and switches by 90°, as substantiated by ab-initio calculations. Our study not only concludes the atomic origin of Tb-Co atom bonding configuration in shaping bulk PMA, but also establishes the groundwork for electrically programmable ferrimagnetic spintronics, such as controlling domain wall motion and programming artificial spin textures.

cond-mat.mtrl-sci↗

arXiv2024

SuperSalt: Equivariant Neural Network Force Fields for Multicomponent Molten Salts System

Molten salts are crucial for clean energy applications, yet exploring their thermophysical properties across diverse chemical space remains challenging. We present the development of a machine learning interatomic potential (MLIP) called SuperSalt, which targets 11-cation chloride melts and captures the essential physics of molten salts with near-DFT accuracy. Using an efficient workflow that integrates systems of one, two, and 11 components, the SuperSalt potential can accurately predict thermophysical properties such as density, bulk modulus, thermal expansion, and heat capacity. Our model is validated across a broad chemical space, demonstrating excellent transferability. We further illustrate how Bayesian optimization combined with SuperSalt can accelerate the discovery of optimal salt compositions with desired properties. This work provides a foundation for future studies that allows easy extensions to more complex systems, such as those containing additional elements. SuperSalt represents a shift towards a more universal, efficient, and accurate modeling of molten salts for advanced energy applications.

cond-mat.mtrl-sci↗

arXiv2024

Off-stoichiometry engineering of the electrical and optical properties of SrNbO$_3$ by oxide molecular beam epitaxy

The highly conducting and transparent inorganic perovskites SrBO$_3$ with V, Nb, Mo, and their mixtures at the B-site have recently attracted the attention of the oxide electronics community as novel alternative transparent conducting oxides. For different applications from solar cells to transparent electronics, it is desirable to tune the optical transmission window in the ultraviolet (UV), visible and infrared (IR) range. The conventional approach is substitutional design at the A- and/or B-site. Here, we suggest a method by engineering the off-stoichiometry of the perovskite, opening new pathways to broaden the range of applications without adding additional elements. For oxide molecular beam epitaxy grown SrNbO$_3$ on GdScO$_3$ substrates, we show that controlled Sr deficiency shifts the plasma edge from about 2 eV in the visible range into the near-infrared region, 1.37 eV (similar to stoichiometric SrVO$_3$). Here, epitaxial growth allows going beyond the limitations of phase stability set by thermodynamics. The suggested approach opens a new design toolbox by including controlled vacancy sites as quasi-substitutional virtual elements.

cond-mat.mtrl-sci↗

arXiv2024

Generative deep learning for the inverse design of materials

In addition to the forward inference of materials properties using machine learning, generative deep learning techniques applied on materials science allow the inverse design of materials, i.e., assessing the composition-processing-(micro-)structure-property relationships in a reversed way. In this review, we focus on the (micro-)structure-property mapping, i.e., crystal structure-intrinsic property and microstructure-extrinsic property, and summarize comprehensively how generative deep learning can be performed. Three key elements, i.e., the construction of latent spaces for both the crystal structures and microstructures, generative learning approaches, and property constraints, are discussed in detail. A perspective is given outlining the challenges of the existing methods in terms of computational resource consumption, data compatibility, and yield of generation.

cond-mat.mtrl-sci↗

arXiv2024

Stable diffusion for the inverse design of microstructures

In materials science, microstructures and their associated extrinsic properties are critical for engineering advanced structural and functional materials, yet their robust reconstruction and generation remain significant challenges. In this work, we developed a microstructure generation model based on the Stable Diffusion (SD) model, training it on a dataset of 576,000 2D synthetic microstructures containing both phase and grain orientation information. This model was applied to a range of tasks, including microstructure reconstruction, interpolation, inpainting, and generation. Experimental results demonstrate that our image-based approach can analyze and generate complex microstructural features with exceptional statistical and morphological fidelity. Additionally, by integrating the ControlNet fine-tuning model, we achieved the inverse design of microstructures based on specific properties. Compared to conventional methods, our approach offers greater accuracy, efficiency, and versatility, showcasing its generative potential in exploring previously uncharted microstructures and paving the way for data-driven development of advanced materials with tailored properties.

cond-mat.mtrl-sci↗

arXiv2024

Kalman filter enhanced Adversarial Bayesian optimization for active sampling in inelastic neutron scattering

Spin waves, or magnons, are fundamental excitations in magnetic materials that provide insights into their dynamic properties and interactions. Magnons are the building blocks of magnonics, which offer promising perspectives for data storage, quantum computing, and communication technologies. These excitations are typically measured through inelastic neutron or x-ray scattering techniques, which involve heavy and time-consuming measurements, data processing, and analysis based on various theoretical models. Here, we introduce a machine learning algorithm that integrates adaptive noise reduction and active learning sampling, which enables the restoration from minimal inelastic neutron scattering point data of spin wave information and the accurate extraction of magnetic parameters, including hidden interactions. Our findings, benchmarked against the magnon spectra of CrSBr, significantly enhance the efficiency and accuracy in addressing complex and noisy experimental measurements. This advancement offers a powerful machine learning tool for research in magnonics and spintronics, which can also be extended to other characterization techniques at large facilities.

cond-mat.mtrl-sci↗

arXiv2024

Paying More Attention to Source Context: Mitigating Unfaithful Translations from Large Language Model

Large language models (LLMs) have showcased impressive multilingual machine translation ability. However, unlike encoder-decoder style models, decoder-only LLMs lack an explicit alignment between source and target contexts. Analyzing contribution scores during generation processes revealed that LLMs can be biased towards previously generated tokens over corresponding source tokens, leading to unfaithful translations. To address this issue, we propose to encourage LLMs to pay more attention to the source context from both source and target perspectives in zeroshot prompting: 1) adjust source context attention weights; 2) suppress irrelevant target prefix influence; Additionally, we propose 3) avoiding over-reliance on the target prefix in instruction tuning. Experimental results from both human-collected unfaithfulness test sets focusing on LLM-generated unfaithful translations and general test sets, verify our methods' effectiveness across multiple language pairs. Further human evaluation shows our method's efficacy in reducing hallucinatory translations and facilitating faithful translation generation.

cs.CL↗

arXiv2024

A matter of performance & criticality: a review of rare-earth-based magnetocaloric intermetallic compounds for hydrogen liquefaction

The low efficiency of conventional liquefaction technologies based on the Joule-Thomson expansion makes liquid hydrogen currently not attractive enough for large-scale energy-related technologies that are important for the transition to a carbon-neutral society. Magnetocaloric hydrogen liquefaction has great potential to achieve higher efficiency and is therefore a crucial enabler for affordable liquid hydrogen. Cost-effective magnetocaloric materials with large magnetic entropy and adiabatic temperature changes in the temperature range of 77 $\sim$ 20 K under commercially practicable magnetic fields are the foundation for the success of magnetocaloric hydrogen liquefaction. Heavy rare-earth-based magnetocaloric intermetallic compounds generally show excellent magnetocaloric performances, but the heavy rare-earth elements (Gd, Tb, Dy, Ho, Er, and Tm) are highly critical in resources. Yttrium and light rare-earth elements (La, Ce, Pr, and Nd) are relatively abundant, but their alloys generally show less excellent magnetocaloric properties. A dilemma appears: higher performance or lower criticality? In this review, we study how cryogenic temperature influences magnetocaloric performance by first reviewing heavy rare-earth-based intermetallic compounds. Next, we look at light rare-earth-based, "mixed" rare-earth-based, and Gd-based intermetallic compounds with the nature of the phase transition order taken into consideration, and summarize ways to resolve the dilemma.

cond-mat.mtrl-sci↗

arXiv2024

The role of Debye temperature in achieving large adiabatic temperature changes at cryogenic temperatures: a case study on $Pr_2In$

The excellent magnetic entropy change ($ΔS_T$) in the temperature range of 20 $\sim$ 77 K due to the first-order phase transition makes $Pr_2In$ an intriguing candidate for magnetocaloric hydrogen liquefaction. As an equally important magnetocaloric parameter, the adiabatic temperature change ($ΔT_{ad}$) of $Pr_2In$ associated with the first-order phase transition has not yet been reported. In this work, the $ΔT_{ad}$ of $Pr_2In$ is obtained from heat capacity measurements: 2 K in fields of 2 T and 4.3 K in fields of 5 T. While demonstrating a $ΔT_{ad}$ that is not as impressive as its remarkable $ΔS_T$, $Pr_2In$ exhibits an unusual low Debye temperature ($T_D$) of around 110 K. Based on these two observations, an approach that combines the mean-field and Debye models is developed to study the correlation between $ΔT_{ad}$ and $T_D$. The role of $T_D$ in achieving large $ΔT_{ad}$ is revealed: materials with higher $T_D$ tend to exhibit larger $ΔT_{ad}$, particularly in the cryogenic temperature range. This discovery explains the absence of an outstanding $ΔT_{ad}$ in $Pr_2In$ and can serve as a tool for designing or searching materials with both a large $ΔS_T$ and a $ΔT_{ad}$.

cond-mat.mtrl-sci↗

arXiv2024

Deep-potential enabled multiscale simulation of gallium nitride devices on boron arsenide cooling substrates

High-efficient heat dissipation plays critical role for high-power-density electronics. Experimental synthesis of ultrahigh thermal conductivity boron arsenide (BAs, 1300 W m-1K-1) cooling substrates into the wide-bandgap semiconductor of gallium nitride (GaN) devices has been realized. However, the lack of systematic analysis on the heat transfer across the BAs-GaN interface hampers the practical applications. In this study, by constructing the accurate and high-efficient machine learning interatomic potentials, we performed multiscale simulations of the BAs-GaN heterostructures. Ultrahigh interfacial thermal conductance (ITC) of 265 MW m-2K-1 is achieved, which lies in the well-matched lattice vibrations of BAs and GaN. Moreover, the competition between grain size and boundary resistance was revealed with size increasing from 1 nm to 100 μm. Such deep-potential equipped multiscale simulations not only promote the practical applications of BAs cooling substrates in electronics, but also offer new approach for designing advanced thermal management systems.

cond-mat.mtrl-sci↗

arXiv2023

Demonstration of a Facile and Efficient Strategy for Yield Stress Determination in Large Amplitude Oscillatory Shear: Algebraic Stress Bifurcation

The large amplitude oscillatory shear (LAOS) has been extensively studied for understanding the rheological responses of yield stress fluids. However, the employed methodology for determining the yield stress remains uncertain albeit the fact that many classical or plausible methods exist in the literature. Along these lines, herein, based on Fourier transform (FT) rheology, stress decomposition, and stress bifurcation, a new straightforward method termed as algebraic stress bifurcation was developed. More specifically, the main goal was to determine the yield stress and investigate the solid-liquid transition of fluids in LAOS. A simple and efficient mathematical framework was established and verified by the KVHB, Saramito, Giesekus models, and FT rheology. The main strength of this approach is that only the data from the stress/strain sweep are required instead of Lissajous curves. Alternative curves based on the first harmonic were constructed to demonstrate the non-critical role of both higher harmonics and phenomenological Lissajous curves in determining yield stress. The determined start and end yield points in the solid-liquid transition were compared with the already existing methods. Furthermore, the resulting solid-liquid transition region was analyzed by FT rheology, stress decomposition, and sequence of the physical process to obtain information on nonlinearity and intracycle/intercycle yielding. Our work provides fruitful insights for explaining and reducing the complexities of the stress bifurcation technique by using an easy-to-understand and implement format. Therefore, a concise theoretical framework was introduced for understanding the concept of yield stress, the intercycle yielding process, and the rational choice of yield stress measurement techniques.

physics.flu-dyn↗