arXiv subjects

Xu He

Publications and source records attributed to Xu He.

At least 55 records · Page 3Linked to original sources

arXiv2025

HuMo: Human-Centric Video Generation via Collaborative Multi-Modal Conditioning

Human-Centric Video Generation (HCVG) methods seek to synthesize human videos from multimodal inputs, including text, image, and audio. Existing methods struggle to effectively coordinate these heterogeneous modalities due to two challenges: the scarcity of training data with paired triplet conditions and the difficulty of collaborating the sub-tasks of subject preservation and audio-visual sync with multimodal inputs. In this work, we present HuMo, a unified HCVG framework for collaborative multimodal control. For the first challenge, we construct a high-quality dataset with diverse and paired text, reference images, and audio. For the second challenge, we propose a two-stage progressive multimodal training paradigm with task-specific strategies. For the subject preservation task, to maintain the prompt following and visual generation abilities of the foundation model, we adopt the minimal-invasive image injection strategy. For the audio-visual sync task, besides the commonly adopted audio cross-attention layer, we propose a focus-by-predicting strategy that implicitly guides the model to associate audio with facial regions. For joint learning of controllabilities across multimodal inputs, building on previously acquired capabilities, we progressively incorporate the audio-visual sync task. During inference, for flexible and fine-grained multimodal control, we design a time-adaptive Classifier-Free Guidance strategy that dynamically adjusts guidance weights across denoising steps. Extensive experimental results demonstrate that HuMo surpasses specialized state-of-the-art methods in sub-tasks, establishing a unified framework for collaborative multimodal-conditioned HCVG. Project Page: https://phantom-video.github.io/HuMo.

cs.CV↗

arXiv2025

White-Box Reasoning: Synergizing LLM Strategy and gm/Id Data for Automated Analog Circuit Design

Analog IC design is a bottleneck due to its reliance on experience and inefficient simulations, as traditional formulas fail in advanced nodes. Applying Large Language Models (LLMs) directly to this problem risks mere "guessing" without engineering principles. We present a "synergistic reasoning" framework that integrates an LLM's strategic reasoning with the physical precision of the gm/Id methodology. By empowering the LLM with gm/Id lookup tables, it becomes a quantitative, data-driven design partner. We validated this on a two-stage op-amp, where our framework enabled the Gemini model to meet all TT corner specs in 5 iterations and extended optimization to all PVT corners. A crucial ablation study proved gm/Id data is key for this efficiency and precision; without it, the LLM is slower and deviates. Compared to a senior engineer's design, our framework achieves quasi-expert quality with an order-of-magnitude improvement in efficiency. This work validates a path for true analog design automation by combining LLM reasoning with scientific circuit design methodologies.

cs.AR↗

arXiv2025

BitsAI-Fix: LLM-Driven Approach for Automated Lint Error Resolution in Practice

As enterprise codebases continue to grow in scale and complexity, the volume of lint errors far exceeds engineers' manual remediation capacity, leading to continuous accumulation of technical debt and hindered development efficiency. This paper presents BitsAI-Fix, an automated lint error remediation workflow based on Large Language Models (LLMs), designed to address this critical challenge in industrial-scale environments. BitsAI-Fix employs tree-sitter for context expansion and generates search-and-replace format patches through specially trained LLMs, followed by lint scan re-verification to output final remediation results. Additionally, our approach introduces an innovative progressive reinforcement learning (RL) training strategy that can automatically acquire verifiable training data during the project cold-start phase and continuously iterate the model by collecting online samples through feedback after system deployment. Furthermore, we designed a targeted rule-based reward mechanism that combines format rewards and correctness rewards while penalizing redundant modifications. We also propose a "code diff matching" methodology to continuously track online effectiveness. In production deployment at ByteDance, our solution has supported over 5,000 engineers, resolved more than 12,000 static analysis issues, achieved approximately 85% remediation accuracy, with around 1,000 weekly active adopters. This work demonstrates the practical feasibility of LLM-based code remediation solutions in enterprise environments and serves as a reference for automated code fix in large-scale industrial scenarios.

cs.SE↗

arXiv2025

Maximin distance designs for mixed continuous, ordinal, and binary variables

Computer experiments are pivotal for modeling complex real-world systems. Maximizing information extraction and ensuring accurate surrogate modeling necessitates space-filling designs, where design points extensively cover the input domain. While substantial research has been conducted on maximin distance designs for continuous variables, which aim to maximize the minimum distance between points, methods accommodating mixed-variable types remain underdeveloped. This paper introduces the first general methodology for constructing maximin distance designs integrating continuous, ordinal, and binary variables. This approach allows flexibility in the number of runs, the mix of variable types, and the granularity of levels for ordinal variables. We propose three advanced algorithms, each rigorously supported by theoretical frameworks, that are computationally efficient and scalable. Our numerical evaluations demonstrate that our methods significantly outperform existing techniques in achieving greater separation distances across design points.

stat.ME↗

arXiv2025

Abinit 2025: New Capabilities for the Predictive Modeling of Solids and Nanomaterials

Abinit is a widely used scientific software package implementing density functional theory and many related functionalities for excited states and response properties. This paper presents the novel features and capabilities, both technical and scientific, which have been implemented over the past 5 years. This evolution occurred in the context of evolving hardware platforms, high-throughput calculation campaigns, and the growing use of machine learning to predict properties based on databases of first principles results. We present new methodologies for ground states with constrained charge, spin or temperature; for density functional perturbation theory extensions to flexoelectricity and polarons; and for excited states in many-body frameworks including GW, dynamical mean field theory, and coupled cluster. Technical advances have extended abinit high-performance execution to graphical processing units and intensive parallelism. Second principles methods build effective models on top of first principles results to scale up in length and time scales. Finally, workflows have been developed in different community frameworks to automate \abinit calculations and enable users to simulate hundreds or thousands of materials in controlled and reproducible conditions.

cond-mat.mtrl-sci↗

arXiv2025

RationalVLA: A Rational Vision-Language-Action Model with Dual System

A fundamental requirement for real-world robotic deployment is the ability to understand and respond to natural language instructions. Existing language-conditioned manipulation tasks typically assume that instructions are perfectly aligned with the environment. This assumption limits robustness and generalization in realistic scenarios where instructions may be ambiguous, irrelevant, or infeasible. To address this problem, we introduce RAtional MAnipulation (RAMA), a new benchmark that challenges models with both unseen executable instructions and defective ones that should be rejected. In RAMA, we construct a dataset with over 14,000 samples, including diverse defective instructions spanning six dimensions: visual, physical, semantic, motion, safety, and out-of-context. We further propose the Rational Vision-Language-Action model (RationalVLA). It is a dual system for robotic arms that integrates the high-level vision-language model with the low-level manipulation policy by introducing learnable latent space embeddings. This design enables RationalVLA to reason over instructions, reject infeasible commands, and execute manipulation effectively. Experiments demonstrate that RationalVLA outperforms state-of-the-art baselines on RAMA by a 14.5% higher success rate and 0.94 average task length, while maintaining competitive performance on standard manipulation tasks. Real-world trials further validate its effectiveness and robustness in practical applications. Our project page is https://irpn-eai.github.io/RationalVLA.

cs.RO↗

arXiv2025

Trajectory World Models for Heterogeneous Environments

Heterogeneity in sensors and actuators across environments poses a significant challenge to building large-scale pre-trained world models on top of this low-dimensional sensor information. In this work, we explore pre-training world models for heterogeneous environments by addressing key transfer barriers in both data diversity and model flexibility. We introduce UniTraj, a unified dataset comprising over one million trajectories from 80 environments, designed to scale data while preserving critical diversity. Additionally, we propose TrajWorld, a novel architecture capable of flexibly handling varying sensor and actuator information and capturing environment dynamics in-context. Pre-training TrajWorld on UniTraj yields substantial gains in transition prediction, achieves a new state-of-the-art for off-policy evaluation, and also delivers superior online performance of model predictive control. To the best of our knowledge, this work, for the first time, demonstrates the transfer benefits of world models across heterogeneous and complex control environments. Code and data are available at https://github.com/thuml/TrajWorld.

cs.LG↗

arXiv2025

Transform-Resampled Double Bootstrap Percentile with Applications in System Reliability Assessment

System reliability assessment(SRA) is a challenging task due to the limited experimental data and the complex nature of the system structures. Despite a long history dating back to \cite{buehler1957confidence}, exact methods have only been applied to SRA for simple systems. High-order asymptotic methods, such as the Cornish-Fisher expansion, have become popular for balancing computational efficiency with improved accuracy when data are limited, but frequently encounter the "bend-back" problem in high-reliability scenarios and require complex analytical computations. To overcome these limitations, we propose a novel method for SRA by modifying the double bootstrap framework, termed the double bootstrap percentile with transformed resamples. In particular, we design a nested resampling process for log-location-scale lifetime models, eliminating the computational burden caused by the iterative resampling process involved in the conventional double bootstrap. We prove that the proposed method maintains the high-order convergence property, thus providing a highly accurate yet computationally efficient confidence limit for system reliability. Moreover, the proposed procedure is straightforward to implement, involving only a simple resampling operation and efficient moment estimation steps. Numerical studies further demonstrate that our approach outperforms the state-of-the-art SRA methods and, at the same time, is much less susceptible to the bend-back issue.

stat.ME↗

arXiv2025

Optimized and regularly repeated lattice-based Latin hypercube designs for large-scale computer experiments

Computer simulations serve as powerful tools for scientists and engineers to gain insights into complex systems. Less costly than physical experiments, computer experiments sometimes involve large number of trials. Conventional design optimization and model fitting methods for computer experiments are inefficient for large-scale problems. In this paper, we propose new methods to optimize good lattice point sets, using less computation to construct designs with enhanced space-filling properties such as high separation distance, low discrepancy, and high separation distance on projections. These designs show promising performance in uncertainty quantification as well as physics-informed neural networks. We also propose a new type of space-filling design called regularly repeated lattice-based Latin hypercube designs, which contain lots of local space-filling Latin hypercube designs as subdesigns. Such designs facilitate rapid fitting of multiple local Gaussian process models in a moving window type of modeling approach and thus are useful for large-scale emulation problems.

stat.ME↗

arXiv2025

Adaptive Grid Designs for Classifying Monotonic Binary Deterministic Computer Simulations

This research is motivated by the need for effective classification in ice-breaking dynamic simulations, aimed at determining the conditions under which an underwater vehicle will break through the ice. This simulation is extremely time-consuming and yields deterministic, binary, and monotonic outcomes. Detecting the critical edge between the negative-outcome and positive-outcome regions with minimal simulation runs necessitates an efficient experimental design for selecting input values. In this paper, we derive lower bounds on the number of functional evaluations needed to ensure a certain level of classification accuracy for arbitrary static and adaptive designs. We also propose a new class of adaptive designs called adaptive grid designs, which are sequences of grids with increasing resolution such that lower resolution grids are proper subsets of higher resolution grids. By prioritizing simulation runs at lower resolution points and skipping redundant runs, adaptive grid designs require the same order of magnitude of runs as the best possible adaptive design, which is an order of magnitude fewer than the best possible static design. Numerical results across test functions, the road crash simulation and the ice-breaking simulation validate the superiority of adaptive grid designs.

stat.ME↗

arXiv2025

SentinelAgent: Graph-based Anomaly Detection in Multi-Agent Systems

The rise of large language model (LLM)-based multi-agent systems (MAS) introduces new security and reliability challenges. While these systems show great promise in decomposing and coordinating complex tasks, they also face multi-faceted risks across prompt manipulation, unsafe tool usage, and emergent agent miscoordination. Existing guardrail mechanisms offer only partial protection, primarily at the input-output level, and fall short in addressing systemic or multi-point failures in MAS. In this work, we present a system-level anomaly detection framework tailored for MAS, integrating structural modeling with runtime behavioral oversight. Our approach consists of two components. First, we propose a graph-based framework that models agent interactions as dynamic execution graphs, enabling semantic anomaly detection at node, edge, and path levels. Second, we introduce a pluggable SentinelAgent, an LLM-powered oversight agent that observes, analyzes, and intervenes in MAS execution based on security policies and contextual reasoning. By bridging abstract detection logic with actionable enforcement, our method detects not only single-point faults and prompt injections but also multi-agent collusion and latent exploit paths. We validate our framework through two case studies, including an email assistant and Microsoft's Magentic-One system, demonstrating its ability to detect covert risks and provide explainable root-cause attribution. Our work lays the foundation for more trustworthy, monitorable, and secure agent-based AI ecosystems.

cs.AI↗

arXiv2025

Pangu Embedded: An Efficient Dual-system LLM Reasoner with Metacognition

This work presents Pangu Embedded, an efficient Large Language Model (LLM) reasoner developed on Ascend Neural Processing Units (NPUs), featuring flexible fast and slow thinking capabilities. Pangu Embedded addresses the significant computational costs and inference latency challenges prevalent in existing reasoning-optimized LLMs. We propose a two-stage training framework for its construction. In Stage 1, the model is finetuned via an iterative distillation process, incorporating inter-iteration model merging to effectively aggregate complementary knowledge. This is followed by reinforcement learning on Ascend clusters, optimized by a latency-tolerant scheduler that combines stale synchronous parallelism with prioritized data queues. The RL process is guided by a Multi-source Adaptive Reward System (MARS), which generates dynamic, task-specific reward signals using deterministic metrics and lightweight LLM evaluators for mathematics, coding, and general problem-solving tasks. Stage 2 introduces a dual-system framework, endowing Pangu Embedded with a "fast" mode for routine queries and a deeper "slow" mode for complex inference. This framework offers both manual mode switching for user control and an automatic, complexity-aware mode selection mechanism that dynamically allocates computational resources to balance latency and reasoning depth. Experimental results on benchmarks including AIME 2024, GPQA, and LiveCodeBench demonstrate that Pangu Embedded with 7B parameters, outperforms similar-size models like Qwen3-8B and GLM4-9B. It delivers rapid responses and state-of-the-art reasoning quality within a single, unified model architecture, highlighting a promising direction for developing powerful yet practically deployable LLM reasoners.

cs.CL↗

arXiv2025

Pangu Ultra MoE: How to Train Your Big MoE on Ascend NPUs

Sparse large language models (LLMs) with Mixture of Experts (MoE) and close to a trillion parameters are dominating the realm of most capable language models. However, the massive model scale poses significant challenges for the underlying software and hardware systems. In this paper, we aim to uncover a recipe to harness such scale on Ascend NPUs. The key goals are better usage of the computing resources under the dynamic sparse model structures and materializing the expected performance gain on the actual hardware. To select model configurations suitable for Ascend NPUs without repeatedly running the expensive experiments, we leverage simulation to compare the trade-off of various model hyperparameters. This study led to Pangu Ultra MoE, a sparse LLM with 718 billion parameters, and we conducted experiments on the model to verify the simulation results. On the system side, we dig into Expert Parallelism to optimize the communication between NPU devices to reduce the synchronization overhead. We also optimize the memory efficiency within the devices to further reduce the parameter and activation management overhead. In the end, we achieve an MFU of 30.0% when training Pangu Ultra MoE, with performance comparable to that of DeepSeek R1, on 6K Ascend NPUs, and demonstrate that the Ascend system is capable of harnessing all the training stages of the state-of-the-art language models. Extensive experiments indicate that our recipe can lead to efficient training of large-scale sparse language models with MoE. We also study the behaviors of such models for future reference.

cs.CL↗

arXiv2025

Switchable Skyrmion-Antiskyrmion Tubes in Rhombohedral BaTiO$_\mathrm{3}$ and Related Materials

Skyrmions are stable topological textures that have garnered substantial attention within the ferroelectric community for their exotic functional properties. While previous studies have questioned the feasibility of [001]$_{\text{pc}}$ skyrmion tubes in rhombohedral BaTiO$_3$ due to the high energy cost of 180$^\circ$ domain walls, we demonstrate here their stabilization with topological charges of $\mathcal{Q} = \pm 1$ from density functional theory and second-principles calculations. By enabling extensive vortex and antivortex polarization configurations, the expected prohibitive energetic barriers are overcomed while preserving the topological nature of the structures. Notably, we extend these findings to demonstrate the appearance of skyrmion and antiskyrmion tubes in other related materials, highlighting their broader relevance. Furthermore, our computational experiments indicate that these structures can be directly stabilized and reversibly switched by applied electric fields, establishing a straightforward route for their practical realization and functional control in nanoelectronic devices.

cond-mat.mtrl-sci↗

arXiv2025

Landmarks Are Alike Yet Distinct: Harnessing Similarity and Individuality for One-Shot Medical Landmark Detection

Landmark detection plays a crucial role in medical imaging applications such as disease diagnosis, bone age estimation, and therapy planning. However, training models for detecting multiple landmarks simultaneously often encounters the "seesaw phenomenon", where improvements in detecting certain landmarks lead to declines in detecting others. Yet, training a separate model for each landmark increases memory usage and computational overhead. To address these challenges, we propose a novel approach based on the belief that "landmarks are distinct" by training models with pseudo-labels and template data updated continuously during the training process, where each model is dedicated to detecting a single landmark to achieve high accuracy. Furthermore, grounded on the belief that "landmarks are also alike", we introduce an adapter-based fusion model, combining shared weights with landmark-specific weights, to efficiently share model parameters while allowing flexible adaptation to individual landmarks. This approach not only significantly reduces memory and computational resource requirements but also effectively mitigates the seesaw phenomenon in multi-landmark training. Experimental results on publicly available medical image datasets demonstrate that the single-landmark models significantly outperform traditional multi-point joint training models in detecting individual landmarks. Although our adapter-based fusion model shows slightly lower performance compared to the combined results of all single-landmark models, it still surpasses the current state-of-the-art methods while achieving a notable improvement in resource efficiency.

cs.CV↗

arXiv2025

Visible-Thermal Tiny Object Detection: A Benchmark Dataset and Baselines

Small object detection (SOD) has been a longstanding yet challenging task for decades, with numerous datasets and algorithms being developed. However, they mainly focus on either visible or thermal modality, while visible-thermal (RGBT) bimodality is rarely explored. Although some RGBT datasets have been developed recently, the insufficient quantity, limited category, misaligned images and large target size cannot provide an impartial benchmark to evaluate multi-category visible-thermal small object detection (RGBT SOD) algorithms. In this paper, we build the first large-scale benchmark with high diversity for RGBT SOD (namely RGBT-Tiny), including 115 paired sequences, 93K frames and 1.2M manual annotations. RGBT-Tiny contains abundant targets (7 categories) and high-diversity scenes (8 types that cover different illumination and density variations). Note that, over 81% of targets are smaller than 16x16, and we provide paired bounding box annotations with tracking ID to offer an extremely challenging benchmark with wide-range applications, such as RGBT fusion, detection and tracking. In addition, we propose a scale adaptive fitness (SAFit) measure that exhibits high robustness on both small and large targets. The proposed SAFit can provide reasonable performance evaluation and promote detection performance. Based on the proposed RGBT-Tiny dataset and SAFit measure, extensive evaluations have been conducted, including 23 recent state-of-the-art algorithms that cover four different types (i.e., visible generic detection, visible SOD, thermal SOD and RGBT object detection). Project is available at https://github.com/XinyiYing/RGBT-Tiny.

cs.CV↗

arXiv2025

TYPEPULSE: Detecting Type Confusion Bugs in Rust Programs

Rust supports type conversions and safe Rust guarantees the security of these conversions through robust static type checking and strict ownership guidelines. However, there are instances where programmers need to use unsafe Rust for certain type conversions, especially those involving pointers. Consequently, these conversions may cause severe memory corruption problems. Despite extensive research on type confusion bugs in C/C++, studies on type confusion bugs in Rust are still lacking. Also, due to Rust's new features in the type system, existing solutions in C/C++ cannot be directly applied to Rust. In this paper, we develop a static analysis tool called TYPEPULSE to detect three main categories of type confusion bugs in Rust including misalignment, inconsistent layout, and mismatched scope. TYPEPULSE first performs a type conversion analysis to collect and determine trait bounds for type pairs. Moreover, it performs a pointer alias analysis to resolve the alias relationship of pointers. Following the integration of information into the property graph, it constructs type patterns and detects each type of bug in various conversion scenarios. We run TYPEPULSE on the top 3,000 Rust packages and uncover 71 new type confusion bugs, exceeding the total number of type confusion bugs reported in RUSTSEC over the past five years. We have received 32 confirmations from developers, along with one CVE ID and six RUSTSEC IDs.

cs.CR↗

arXiv2025

Stacking-engineered ferroelectricity and multiferroic order in van der Waals magnets

Two-dimensional (2D) materials that exhibit spontaneous magnetization, polarization or strain (referred to as ferroics) have the potential to revolutionize nanotechnology by enhancing the multifunctionality of nanoscale devices. However, multiferroic order is difficult to achieve, requiring complicated coupling between electron and spin degrees of freedom. We propose a universal method to engineer multiferroics from van der Waals magnets by taking advantage of the fact that changing the stacking between 2D layers can break inversion symmetry, resulting in ferroelectricity and possibly magnetoelectric coupling. We illustrate this concept using first-principles calculations in bilayer NiI$_2$, which can be made ferroelectric upon rotating two adjacent layers by $180^{\circ}$ with respect to the bulk stacking. Furthermore, we discover a novel multiferroic order induced by interlayer charge transfer which couples the interlayer spin order and electronic polarization. Our approach is not only general but also systematic, and can enable the discovery of a wide variety of 2D multiferroics.

cond-mat.mtrl-sci↗