arXiv subjects

Xiwen Zhang

Publications and source records attributed to Xiwen Zhang.

At least 19 recordsLinked to original sources

arXiv2026

On Demand magnetic-Doppler nuclear frequency comb memory for hard X-ray photons

Nuclear quantum memories in the hard X-ray regime offer some key advantages over their optical counterparts, such as broader bandwidth and lower background noise. A Doppler frequency comb protocol has been theoretically proposed [X. Zhang \textit{et al.}, Phys. Rev. Lett. \textbf{123}, 250504 (2019)] and recently demonstrated experimentally [S. Velten \textit{et al.}, Sci. Adv. \textbf{10}, eadn9825 (2024)] for the storage and retrieval of X-ray photons. However, achieving on-demand retrieval remains challenging because of the requirement for precise and synchronous mechanical motion of multiple absorbers. We propose a hybrid, magnetic-Doppler nuclear frequency comb composed of Doppler-shifted resonant absorbers with lifted nuclear spin degeneracy, which expands the Doppler comb structure. By synchronously reversing the directions of both the magnetic fields and absorber velocities, the system achieves time-reversed phase evolution dynamics that allows for efficient on-demand photon retrieval with significantly reduced mechanical complexity.

quant-ph↗

arXiv2026

Electrical manipulation and detection of perpendicular altermagnetic order via proximitized Dirac semimetal

Altermagnets, which combine antiferromagnetic-like magnetic compensation with ferromagnetic-like broken time-reversal symmetry, hold great promise for high-density and ultrafast spintronic applications. However, the detection and switching of perpendicular altermagnetic order are fundamentally constrained by magnetic symmetry, restricting both fundamental studies and practical implementation. We realize robust electrical reading and deterministic switching of perpendicular altermagnetic order by designing a Dirac semimetal/altermagnet heterostructure of PtTe2/CrSb. This engineered interface enables anomalous Hall readout via altermagnetic proximity effect and delivers efficient spin-orbit torque for manipulating the epitaxial perpendicular Neel vector in CrSb. These findings significantly broaden the functional scope of altermagnetic heterostructures and pave the way for highly scalable altermagnetic memory.

cond-mat.mtrl-sci↗

arXiv2025

On Structured State-Space Duality

Structured State-Space Duality (SSD) [Dao & Gu, ICML 2024] is an equivalence between a simple Structured State-Space Model (SSM) and a masked attention mechanism. In particular, a state-space model with a scalar-times-identity state matrix is equivalent to a masked self-attention with a $1$-semiseparable causal mask. Consequently, the same sequence transformation (model) has two algorithmic realizations: as a linear-time $O(T)$ recurrence or as a quadratic-time $O(T^2)$ attention. In this note, we formalize and generalize this duality: (i) we extend SSD from the scalar-identity case to general diagonal SSMs (diagonal state matrices); (ii) we show that these diagonal SSMs match the scalar case's training complexity lower bounds while supporting richer dynamics; (iii) we establish a necessary and sufficient condition under which an SSM is equivalent to $1$-semiseparable masked attention; and (iv) we show that such duality fails to extend to standard softmax attention due to rank explosion. Together, these results tighten bridge between recurrent SSMs and Transformers, and widen the design space for expressive yet efficient sequence models.

cs.LG↗

arXiv2025

Co-TAP: Three-Layer Agent Interaction Protocol Technical Report

This paper proposes Co-TAP (T: Triple, A: Agent, P: Protocol), a three-layer agent interaction protocol designed to address the challenges faced by multi-agent systems across the three core dimensions of Interoperability, Interaction and Collaboration, and Knowledge Sharing. We have designed and proposed a layered solution composed of three core protocols: the Human-Agent Interaction Protocol (HAI), the Unified Agent Protocol (UAP), and the Memory-Extraction-Knowledge Protocol (MEK). HAI focuses on the interaction layer, standardizing the flow of information between users, interfaces, and agents by defining a standardized, event-driven communication paradigm. This ensures the real-time performance, reliability, and synergy of interactions. As the core of the infrastructure layer, UAP is designed to break down communication barriers among heterogeneous agents through unified service discovery and protocol conversion mechanisms, thereby enabling seamless interconnection and interoperability of the underlying network. MEK, in turn, operates at the cognitive layer. By establishing a standardized ''Memory (M) - Extraction (E) - Knowledge (K)'' cognitive chain, it empowers agents with the ability to learn from individual experiences and form shareable knowledge, thereby laying the foundation for the realization of true collective intelligence. We believe this protocol framework will provide a solid engineering foundation and theoretical guidance for building the next generation of efficient, scalable, and intelligent multi-agent applications.

cs.AI↗

arXiv2025

On the Perception Bottleneck of VLMs for Chart Understanding

Chart understanding requires models to effectively analyze and reason about numerical data, textual elements, and complex visual components. Our observations reveal that the perception capabilities of existing large vision-language models (LVLMs) constitute a critical bottleneck in this process. In this study, we delve into this perception bottleneck by decomposing it into two components: the vision encoder bottleneck, where the visual representation may fail to encapsulate the correct information, and the extraction bottleneck, where the language model struggles to extract the necessary information from the provided visual representations. Through comprehensive experiments, we find that (1) the information embedded within visual representations is substantially richer than what is typically captured by linear extractors, such as the widely used retrieval accuracy metric; (2) While instruction tuning effectively enhances the extraction capability of LVLMs, the vision encoder remains a critical bottleneck, demanding focused attention and improvement. Therefore, we further enhance the visual encoder to mitigate the vision encoder bottleneck under a contrastive learning framework. Empirical results demonstrate that our approach significantly mitigates the perception bottleneck and improves the ability of LVLMs to comprehend charts. Code is publicly available at https://github.com/hkust-nlp/Vision4Chart.

cs.CV↗

arXiv2025

Probing the Linewidth of the 12.4-keV Solid-State $^{45}$Sc Isomeric Resonance

The $^{45}$Sc nuclear transition from the ground to the isomeric state at 12.389~keV, with a lifetime of 0.46~s, exhibits an extraordinarily narrow natural width of 1.4~feV and a quality factor $\simeq 10^{19}$ -- surpassing those of the most precise atomic clocks -- making $^{45}$Sc a compelling platform for advanced metrology and nuclear clocks. Here we investigate how closely the spectral width and quality factor of the solid-state $^{45}$Sc resonance can approach these natural limits. Using the European X-ray Free-Electron Laser, we confirm the isomer's lifetime via time-delayed incoherent $K_{α,β}$ fluorescence and observe previously unreported elastic fluorescence, yielding a partial internal conversion coefficient of 390(60). The absence of a clear nuclear forward scattering signal beyond a 2-ms delay implies environmental broadening of at least $500~Γ_{0}$ under experimental conditions, placing bounds on solid-state decoherence mechanisms. These findings set new experimental benchmarks for solid-state nuclear clock development.

quant-ph↗

arXiv2025

On-Demand Zeeman Nuclear Frequency Comb Quantum Memory

The emerging hard X-ray - nuclear interfaces offer unique potential advantages over traditional optical-atomic interfaces for room-temperature, solid-state quantum information processing, including lower background noise, tighter focusing, and exceptionally high resonance quality. Leveraging such interfaces, a major milestone was recently achieved with the first implementation of nuclear quantum memory in the hard X-ray range [S. Velten et al., Nuclear quantum memory for hard X-ray photon wave packets, Sci. Adv. 10, eadn9825 (2024)] using the Doppler frequency comb protocol. However, this approach relies on the synchronous mechanical motion of multiple nuclear absorbers, posing experimental challenges for on-demand photon retrieval. We propose an on-demand hard X-ray quantum memory based on reversing the direction of an external magnetic field in a single stationary solid-state nuclear absorber with sets of Zeeman sublevels. This scheme is exemplified by the quantum storage of an 1.41-$μ$s single photon wave packet at 6.2 keV for over 10 $μ$s in a $^{181}$Ta metallic foil, providing a feasible pathway for the first experimental demonstration of on-demand hard X-ray photon storage.

quant-ph↗

arXiv2025

Extreme Cardiac MRI Analysis under Respiratory Motion: Results of the CMRxMotion Challenge

Deep learning models have achieved state-of-the-art performance in automated Cardiac Magnetic Resonance (CMR) analysis. However, the efficacy of these models is highly dependent on the availability of high-quality, artifact-free images. In clinical practice, CMR acquisitions are frequently degraded by respiratory motion, yet the robustness of deep learning models against such artifacts remains an underexplored problem. To promote research in this domain, we organized the MICCAI CMRxMotion challenge. We curated and publicly released a dataset of 320 CMR cine series from 40 healthy volunteers who performed specific breathing protocols to induce a controlled spectrum of motion artifacts. The challenge comprised two tasks: 1) automated image quality assessment to classify images based on motion severity, and 2) robust myocardial segmentation in the presence of motion artifacts. A total of 22 algorithms were submitted and evaluated on the two designated tasks. This paper presents a comprehensive overview of the challenge design and dataset, reports the evaluation results for the top-performing methods, and further investigates the impact of motion artifacts on five clinically relevant biomarkers. All resources and code are publicly available at: https://github.com/CMRxMotion

eess.IV↗

arXiv2025

Diving into Self-Evolving Training for Multimodal Reasoning

Self-evolving trainin--where models iteratively learn from their own outputs--has emerged as a key approach for complex reasoning tasks, addressing the scarcity of high-quality chain-of-thought data. However, its effectiveness in multimodal reasoning, a domain more intricate than text-only reasoning, remains underexplored, and the understanding of critical factors in this training paradigm remains limited. Furthermore, a central challenge for this training method is performance saturation, which impedes further improvements and scalability. Inspired by reinforcement learning (RL), in this paper, we reframe self-evolving training for multimodal reasoning through the lens of RL, identifying three pivotal factors: Training Method, Reward Model, and Prompt Variation. Through systematic analysis, we establish relatively optimal design principles that significantly enhance multimodal reasoning capabilities. Moreover, delving deeper into training dynamics, we uncover the roots of saturation and propose a new automatic balancing mechanism to mitigate this limitation. Building on these insights, we propose M-STAR (Multimodal Self-evolving Training for Reasoning), a framework that achieves consistent performance gains across models of varying sizes and diverse benchmarks. All resources are made publicly available at https://mstar-lmm.github.io.

cs.CL↗

arXiv2025

Minimalist Softmax Attention Provably Learns Constrained Boolean Functions

We study the computational limits of learning $k$-bit Boolean functions (specifically, $\mathrm{AND}$, $\mathrm{OR}$, and their noisy variants), using a minimalist single-head softmax-attention mechanism, where $k=Θ(d)$ relevant bits are selected from $d$ inputs. We show that these simple $\mathrm{AND}$ and $\mathrm{OR}$ functions are unsolvable with a single-head softmax-attention mechanism alone. However, with teacher forcing, the same minimalist attention is capable of solving them. These findings offer two key insights: Architecturally, solving these Boolean tasks requires only minimalist attention, without deep Transformer blocks or FFNs. Methodologically, one gradient descent update with supervision suffices and replaces the multi-step Chain-of-Thought (CoT) reasoning scheme of [Kim and Suzuki, ICLR 2025] for solving Boolean problems. Together, the bounds expose a fundamental gap between what this minimal architecture achieves under ideal supervision and what is provably impossible under standard training.

cs.LG↗

arXiv2025

A LSTM-Transformer Model for pulsation control of pVADs

Methods: A method of the pulsation for a pVAD is proposed (AP-pVAD Model). AP-pVAD Model consists of two parts: NPQ Model and LSTM-Transformer Model. (1)The NPQ Model determines the mathematical relationship between motor speed, pressure, and flow rate for the pVAD. (2)The Attention module of Transformer neural network is integrated into the LSTM neural network to form the new LSTM-Transformer Model to predict the pulsation time characteristic points for adjusting the motor speed of the pVAD. Results: The AP-pVAD Model is validated in three hydraulic experiments and an animal experiment. (1)The pressure provided by pVAD calculated with the NPQ Model has a maximum error of only 2.15 mmHg compared to the expected values. (2)The pulsation time characteristic points predicted by the LSTM-Transformer Model shows a maximum prediction error of 1.78ms, which is significantly lower than other methods. (3)The in-vivo test of pVAD in animal experiment has significant improvements in aortic pressure. Animals survive for over 27 hours after the initiation of pVAD operation. Conclusion: (1)For a given pVAD, motor speed has a linear relationship with pressure and a quadratic relationship with flow. (2)Deep learning can be used to predict pulsation characteristic time points, with the LSTM-Transformer Model demonstrating minimal prediction error and better robust performance under conditions of limited dataset sizes, elevated noise levels, and diverse hyperparameter combinations, demonstrating its feasibility and effectiveness.

physics.med-ph↗

arXiv2025

Small Language Models: Survey, Measurements, and Insights

Small language models (SLMs), despite their widespread adoption in modern smart devices, have received significantly less academic attention compared to their large language model (LLM) counterparts, which are predominantly deployed in data centers and cloud environments. While researchers continue to improve the capabilities of LLMs in the pursuit of artificial general intelligence, SLM research aims to make machine intelligence more accessible, affordable, and efficient for everyday tasks. Focusing on transformer-based, decoder-only language models with 100M-5B parameters, we survey 70 state-of-the-art open-source SLMs, analyzing their technical innovations across three axes: architectures, training datasets, and training algorithms. In addition, we evaluate their capabilities in various domains, including commonsense reasoning, mathematics, in-context learning, and long context. To gain further insight into their on-device runtime costs, we benchmark their inference latency and memory footprints. Through in-depth analysis of our benchmarking data, we offer valuable insights to advance research in this field.

cs.CL↗

arXiv2024

DART-Math: Difficulty-Aware Rejection Tuning for Mathematical Problem-Solving

Solving mathematical problems requires advanced reasoning abilities and presents notable challenges for large language models. Previous works usually synthesize data from proprietary models to augment existing datasets, followed by instruction tuning to achieve top-tier results. However, our analysis of these datasets reveals severe biases towards easy queries, with frequent failures to generate any correct response for the most challenging queries. Hypothesizing that difficult queries are crucial to learn complex reasoning, we propose Difficulty-Aware Rejection Tuning (DART), a method that allocates difficult queries more trials during the synthesis phase, enabling more extensive training on difficult samples. Utilizing DART, we have created new datasets for mathematical problem-solving that focus more on difficult queries and are substantially smaller than previous ones. Remarkably, our synthesis process solely relies on a 7B-sized open-weight model, without reliance on the commonly used proprietary GPT-4. We fine-tune various base models on our datasets ranging from 7B to 70B in size, resulting in a series of strong models called DART-MATH. In comprehensive in-domain and out-of-domain evaluation on 6 mathematical benchmarks, DART-MATH outperforms vanilla rejection tuning significantly, being superior or comparable to previous arts, despite using much smaller datasets and no proprietary models. Furthermore, our results position our synthetic datasets as the most effective and cost-efficient publicly available resources for advancing mathematical problem-solving.

cs.CL↗

arXiv2024

PhoneLM:an Efficient and Capable Small Language Model Family through Principled Pre-training

The interest in developing small language models (SLM) for on-device deployment is fast growing. However, the existing SLM design hardly considers the device hardware characteristics. Instead, this work presents a simple yet effective principle for SLM design: architecture searching for (near-)optimal runtime efficiency before pre-training. Guided by this principle, we develop PhoneLM SLM family (currently with 0.5B and 1.5B versions), that acheive the state-of-the-art capability-efficiency tradeoff among those with similar parameter size. We fully open-source the code, weights, and training datasets of PhoneLM for reproducibility and transparency, including both base and instructed versions. We also release a finetuned version of PhoneLM capable of accurate Android Intent invocation, and an end-to-end Android demo. All materials are available at https://github.com/UbiquitousLearning/PhoneLM.

cs.CL↗

arXiv2024

Evaluation of Vortex Criteria by Virtue of the Quadruple Decomposition of Velocity Gradient Tensor

Based on the analysis of the velocity gradient tensor, we investigate in this paper the physical interpretation and limitations of four vortex criteria: $ω$, $Q$, $\varDelta$ and $λ_{ci}$, and reveal the actual physical meaning of vortex patterns which are usually illustrated by level sets of various vortex criteria. A quadruple decomposition based on the normality of the velocity gradient tensor is proposed for the first time, which resolves the motion of a fluid element into dilation, axial stretch along the normal frame, in-plane distortion, and simple shear, in order to clarify the kinematical interpretation of various vortex criteria. The mean rotation characterized by the vorticity $ω$ always consists of simple shear; the $Q$-criterion can reflect the strength of net rotation within the invariant plane relative to the axial stretch of a fluid element, and it is a sufficient but unnecessary condition for the existence of net rotation; the $\varDelta$-criterion can exactly identify the existence of net rotation, but it is not the strength of net rotation; in the case that net rotation exists, $λ_{ci}$ characterize its absolute strength. net rotation is the total effect of the normal rotation within the invariant plane and the simple shear, where the former is the most basic rotation. The newly introduced quadruple decomposition can improve our understanding of vortices in fluids and their motions.

physics.flu-dyn↗

arXiv2024

InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation

Diffusion models have revolutionized text-to-image generation with its exceptional quality and creativity. However, its multi-step sampling process is known to be slow, often requiring tens of inference steps to obtain satisfactory results. Previous attempts to improve its sampling speed and reduce computational costs through distillation have been unsuccessful in achieving a functional one-step model. In this paper, we explore a recent method called Rectified Flow, which, thus far, has only been applied to small datasets. The core of Rectified Flow lies in its \emph{reflow} procedure, which straightens the trajectories of probability flows, refines the coupling between noises and images, and facilitates the distillation process with student models. We propose a novel text-conditioned pipeline to turn Stable Diffusion (SD) into an ultra-fast one-step model, in which we find reflow plays a critical role in improving the assignment between noise and images. Leveraging our new pipeline, we create, to the best of our knowledge, the first one-step diffusion-based text-to-image generator with SD-level image quality, achieving an FID (Frechet Inception Distance) of $23.3$ on MS COCO 2017-5k, surpassing the previous state-of-the-art technique, progressive distillation, by a significant margin ($37.2$ $\rightarrow$ $23.3$ in FID). By utilizing an expanded network with 1.7B parameters, we further improve the FID to $22.4$. We call our one-step models \emph{InstaFlow}. On MS COCO 2014-30k, InstaFlow yields an FID of $13.1$ in just $0.09$ second, the best in $\leq 0.1$ second regime, outperforming the recent StyleGAN-T ($13.9$ in $0.1$ second). Notably, the training of InstaFlow only costs 199 A100 GPU days. Codes and pre-trained models are available at \url{github.com/gnobitab/InstaFlow}.

cs.LG↗

arXiv2023

FastFold: Reducing AlphaFold Training Time from 11 Days to 67 Hours

Protein structure prediction helps to understand gene translation and protein function, which is of growing interest and importance in structural biology. The AlphaFold model, which used transformer architecture to achieve atomic-level accuracy in protein structure prediction, was a significant breakthrough. However, training and inference of the AlphaFold model are challenging due to its high computation and memory cost. In this work, we present FastFold, an efficient implementation of AlphaFold for both training and inference. We propose Dynamic Axial Parallelism and Duality Async Operations to improve the scaling efficiency of model parallelism. Besides, AutoChunk is proposed to reduce memory cost by over 80% during inference by automatically determining the chunk strategy. Experimental results show that FastFold reduces overall training time from 11 days to 67 hours and achieves 7.5X - 9.5X speedup for long-sequence inference. Furthermore, we scale FastFold to 512 GPUs and achieve an aggregate throughput of 6.02 PetaFLOP/s with 90.1% parallel efficiency.

cs.LG↗

arXiv2022

Structural reconstruction and anisotropic conductance in $4f$-ferromagnetic monolayer

Two-dimensional magnets are promising for nanoscale spintronic applications. Currently, most available candidates are based on $3d$ transition metal compounds, with hexagonal or honeycomb lattice geometry. Here, a GdCl$_3$ monolayer with $4f$ moments is theoretically studied, which can be exfoliated from its existing bulk. Its orthorhombic structure and hendecahedral ion cages are unique in two-dimensional. Furthermore, a significant structural reconstruction is caused by the implantation of Li atoms into its interstitial position, which also lead to ferromagnetism via a double-exchange-like process. Its highly anisotropic conductance may be peculiarly useful for nanoelectronics.

cond-mat.mtrl-sci↗