arXiv subjects

Bo Yang

Publications and source records attributed to Bo Yang.

At least 19 recordsLinked to original sources

arXiv2026

Learning Interaction between Image and Layout Priors for Joint Image-Layout Generation in Design Templates

In this paper, we address the problem of graphic design template creation, which generates a background image and a layout of foreground elements over the background to form a harmonious composition from an input text. Prior work on graphic design generation mostly adopts a sequential paradigm, where design elements are generated sequentially. We argue that such a sequential scheme falls short of faithfully capturing the dependency between the background and layout (and thus the joint image-layout distribution), which limits the quality of generated design templates. To overcome this limitation, we propose a model, InterIL, which jointly generates the two modalities, background image and layout, in a single generative process. The novel design of our joint model connects the backbones of pretrained image and layout diffusion models with a learnable communication module to explicitly model bidirectional image-layout interaction. During training, the image and layout backbones are frozen to maintain and leverage the vast pretrained single-modality prior knowledge, while only the communication module is updated, so that the model can focus on learning image-layout interaction and thereby better capture the joint image-layout distribution for improved composition harmony. Our model has no design-specific inductive bias, which allows it to better preserve the original characteristics of realistic designs. We further introduce a test-time guidance strategy to enable users to impose their specific preferences on generated results. Our experiments show that, compared with prior approaches, our model can generate significantly better results in terms of image, layout and image-layout harmonization, producing outputs closer to real samples. We also demonstrate the flexibility of our model in enforcing user preferences at inference without retraining.

cs.CV↗

arXiv2026

SenseNova-U1.5: Towards Native Unified Visual Intelligence

We launch SenseNova-U1.5, an 8B-MoT native unified multimodal model that understands, reasons about, and generates visual content within an encoder-free and VAE-free architecture. We strengthen its visual interface through spatially coherent patch reconstruction and scale its training with carefully curated generation and editing data, improved task formulation, structural prompt enhancement, and native resolutions of up to 4K. For post-training, we optimize specialized experts for visual aesthetics, bilingual text rendering, infographic generation, and image editing, and consolidate their capabilities through multi-expert on-policy distillation. Across extensive evaluations, SenseNova-U1.5 largely advances image fidelity, text rendering, complex composition, multi-reference editing, and interleaved generation, while improving instruction following and preserving subject identity, geometry, and unmodified regions. Despite limited exposure to structured formats in its generation data, SenseNova-U1.5 generalizes effectively to long, complex, and structured visual instructions, further proving that multimodal understanding can transfer to visual planning and creation. Together, these findings position native unified modelling as a promising path towards systems that perceive, reason and create within a fully end-to-end framework. We will open-source training code, including supervised fine-tuning, reinforcement learning, and on-policy distillation.

cs.CV↗

arXiv2026

A Trust-Network-Based Federated Learning Framework for Multi-Center Aging Clock Prediction

Aging clocks quantify biological aging and help characterize individual health status. What protein interactions are important for accurate aging clocks, and are they zeroth-order or higher-order? Addressing these questions requires learning from large molecular datasets distributed across medical centers, where privacy constraints prevent centralized data sharing. Federated learning offers a natural solution but faces four challenges in this setting: limited local sample sizes, sparse and directional inter-center trust, the need to retain discriminative age prediction while supporting interpretation, and model drift and forgetting under heterogeneous cross-center data. We propose TNFL, a trust-network-based federated learning framework that progressively propagates models along directed pairwise trust relations without centralized aggregation. TNFL combines an age-aware mixture-of-experts model with generative replay to preserve previously learned information and reduce forgetting and drift. Experiments across multiple molecular datasets show that TNFL enables effective aging-clock prediction with limited local data, provides interpretable age-dependent prediction patterns, and maintains stable performance across interaction orders. To investigate the biological questions, we analyze TNFL-identified pairwise protein interactions and their higher-order organization through functional and network analyses. The identified interactions repeatedly form coordinated higher-order subnetworks spanning multiple aging-related biological systems, with several proteins recurring across subnetworks. These findings suggest that TNFL captures molecular relationships beyond isolated pairwise associations and reveals coherent higher-order biological organization associated with aging.

cs.LG↗

arXiv2026

Holomorphic functions on complete Hermitian manifolds with flat Chern connection, II

In this paper, we prove several results concerning the function theory of complete Hermitian manifolds with vanishing Chern curvature, which we refer to as complete Chern-flat manifolds. First, we prove that any complete simply connected Chern-flat manifold is elliptic in the sense of Gromov. Moreover, it is biholomorphic to a complex Lie group if the torsion has polynomial growth. Next, we obtain sharp dimension estimates for holomorphic functions of polynomial growth on complete Chern-flat manifolds, with no growth assumption on the torsion. Finally, we prove that any complete Chern-flat manifold admitting polynomial-growth holomorphic functions that form a local holomorphic coordinate system at some point is biholomorphic to complex Euclidean space. These results generalize our previous work, in which sublinear growth of the torsion was assumed. The new analytic tool is a version of Cauchy estimates along orbits of holomorphic flows generated by parallel unitary frames on such manifolds.

math.DG↗

arXiv2026

Tools-CC-Bench: a Benchmark Suite for Collective Communication with Compression in HPC and AI Workloads

Distributed HPC and LLM workloads increasingly require efficient communication for scalability, yet growing data movement has become a major performance bottleneck. Communication compression can reduce this overhead and complement execution-level optimizations, but its benefits remain difficult to assess because existing benchmarks lack support for diverse backends, realistic datasets, application-specific accuracy metrics, and overlap-induced resource contention. We present CC-Bench, a lightweight, extensible, and application-oriented benchmark suite for evaluating communication compression under realistic execution conditions. CC-Bench uses declarative application-environment modeling to decouple profiling logic from communication libraries, datasets, and fidelity metrics, enabling portable cross-library evaluation. It further combines function-level interception and hardware counter monitoring to characterize per-phase latency, hardware utilization, numerical fidelity, and computation interference. With representative datasets from HPC and LLM workloads, CC-Bench evaluates three compression-enabled communication libraries on CPU and GPU clusters, revealing accuracy-performance trade-offs and bottlenecks to guide practical deployment and optimization.

cs.DC↗

arXiv2026

RelightFormer: Feed-forward Generative Transformer for Multiview Object Relighting

Image relighting is traditionally tackled via complex inverse rendering pipelines, which suffer from ill-posed optimization, or single-image generative models that ignore crucial multi-view cues necessary for understanding 3D geometry and material interactions. To address these limitations, we introduce a feed-forward generative Transformer for direct single- and multi-view image relighting that entirely bypasses explicit intrinsic property estimation. Adapted from a video foundation model, our architecture features a latent illumination module that dynamically injects target environment maps into spatial features via cross-attention. Furthermore, we employ permutation-invariant positional encodings to symmetrically process unordered multi-view inputs without sequential bias. To train this robust data-driven model, we construct the massive Laval Objaverse Dataset (LOD), comprising 90K objects and 39K unique illuminations. Extensive experiments demonstrate state-of-the-art visual quality, photorealistic relighting quality, and strong zero-shot generalization across single-view, multi-view, and novel-view relighting tasks.

cs.CV↗

arXiv2026

ReliableRAG: Combating Misinformation in Retrieval-Augmented Generation via Reliability-Guided Reasoning Chains

Retrieval-Augmented Generation (RAG) has emerged as a powerful architecture for Question Answering (QA) by integrating external information into Large Language Models (LLMs). However, false, inaccurate, and misleading information in news and social media poses a serious challenge to real-world RAG systems, especially in multi-hop QA, where complex multi-step reasoning can be misled by even a single deceptive misinformation segment in the retrieved documents. Existing approaches mainly rely on implicit alignment or explicit regulation, but their limited ability to assess fine-grained information reliability makes them vulnerable to deceptive misinformation that is semantically relevant to the question yet factually incorrect, leading to erroneous answers. To address this limitation, we propose ReliableRAG, which, to the best of our knowledge, is the first reliability-driven framework that mitigates deceptive misinformation in multi-hop QA through fine-grained evaluation of individual triples. ReliableRAG first extracts information segments from source documents and represents them as structured triples. It then quantifies triple reliability by combining query-triple semantic relevance with triple credibility, retaining only the top-$K$ reliable and non-redundant triples. Based on these refined triples, ReliableRAG autoregressively constructs robust reasoning chains to consolidate trustworthy evidence and filter deceptive misinformation, producing accurate answers faithful to reliable information. Experiments on three multi-hop QA datasets show that ReliableRAG outperforms existing methods, substantially improving the factual reliability and robustness of RAG systems under deceptive misinformation injection.

cs.CL↗

arXiv2026

Sequential Magnetic Reconnections in a Fishbone-like Structure Leading to Recurrent Brightenings

The fine-scale release of magnetic free energy in the solar atmosphere is a fundamental open question in solar physics. Multi-wavelength observations at high spatiotemporal resolution now offer a direct window into this process. Using data from NVST, SDO, IRIS, and Hinode, we reveal the energy release process in a fishbone-like magnetic structure within active region 12297. The fishbone-like structure consists of a spine along a narrow, elongated positive-polarity field, with herringbone branches rooted in negative-polarity sunspots. Persistent photospheric magnetic flux emergence and shearing motions are observed beneath the fishbone-like structure, which may play a key role in maintaining its topology and producing the recurrent brightenings. During brightenings, compact bright features propagate sequentially from west to east along the spine, accompanied by bidirectional flows along the branches and plasma blobs ejected toward the distant positive sunspot. These propagating features could be interpreted as signatures of sequential magnetic reconnection events occurring in chronological order at nodes along the spine, predominantly in the chromosphere and transition region. Our observations may provide evidence that recurrent brightenings could arise from repeated sequential reconnection events organized by a coherent magnetic structure, deepening our understanding of how such recurrent brightenings are generated and how magnetic free energy is dissipated at fine scales in solar active regions.

astro-ph.SR↗

arXiv2026

Spatiotemporal organization in bike-sharing systems using gravity model parameters

Gravity models describe bike-sharing origin-destination (OD) flows as increasing with origin and destination activity and decreasing with distance. Yet how these relationships change within a day and across observation areas remains unclear. We examine temporal and spatial variation in origin, destination and distance exponents $\alpha$, $\beta$, and $\gamma$, together with $R^2$ and mean absolute error (MAE), across eight bike-sharing systems. Temporal modeling uses sliding windows, while spatial modeling expands a circular area and separates intra-zonal, cross-zonal outflow, cross-zonal inflow and extra-zonal trips. We find recurring patterns across cities: morning and evening peaks differ in origin and destination dependence, while cross-zonal outflow and inflow across the same boundary show opposite changes in their relative origin and destination dependence as radius increases. Model performance and distance dependence also vary with time and radius. A full-day, citywide fit therefore combines time periods and flow types with different gravity relationships.

physics.soc-ph↗

arXiv2026

Scalar curvature of blow-ups of compact K\"ahler manifolds along complex submanifolds

Let $(M,\omega)$ be a compact K\"ahler manifold with its scalar curvature $S(\omega)$, Assume that $M$ has complex dimension at least $3$ and contains a complex submanifold $X$ of complex codimension at least $2$. Let $\sigma: Bl_{X} M \rightarrow M$ denote the blow-up of $M$ along $X$. We show that $Bl_{X} M$ admits a sequence of K\"ahler metrics $\{\widetilde{\omega}_{i}\}_{i \geq 1}$ whose scalar curvatures $S(\widetilde{\omega}_i)$ converge to $\sigma^{\ast} (S(\omega))$ in the $C^0(Bl_{X} M)$ norm. Our work is motivated by a recent result of Brown, who established the corresponding result for blow-ups at a point. The proof is based on the gluing method for constructing extremal K\"ahler metrics on blow-ups, together with new analytic tools and several modifications needed in our setting.

math.DG↗

arXiv2026

Many-Anyon Braiding in Non-Abelian Fractional Quantum Hall Effect with Hybrid Monte Carlo Simulation

We employ the hybrid Monte Carlo method to efficiently compute the many-anyon non-Abelian braiding matrices associated with different braiding schemes of the Moore-Read quasiholes. A novel proposal in this work is that anyon braiding schemes based on a global rotation are robust against finite-size effects, as demonstrated by benchmarking their errors in the braiding matrix against those of a simple two-anyon exchange. Moreover, we investigate how electron-electron interactions and local electrostatic trapping potentials influence the energetic preference of different fusion channels. Their effect on the non-Abelian braiding matrices has been verified, a surprising phenomenon that demonstrates long-range entanglement of non-Abelian states. Our results are relevant to the experimental realization of non-Abelian physics in fractional quantum Hall and other analogous systems, including the fast-growing field of fractional quantum anomalous Hall states in moir\'e materials.

cond-mat.str-el↗

arXiv2026

Breaking the Curse of Multilinguality in Many-to-Many Speech-to-Text Translation via a Resource-Aware Mixture of Speech Encoders

Multimodal large language models (MLLMs) have achieved significant success in speech-to-text translation (S2TT). However, when processing multilingual speech inputs, a single speech encoder shared across all languages suffers from the curse of multilinguality: languages at different resource levels compete for limited representation capacity, leading to strong high-resource performance but substantial degradation on low-resource speech. To address this problem and improve multilingual consistency, we propose MSRT, a novel framework built around a resource-aware Mixture of Speech Encoders (MoSE). MoSE uses an explicit language router to assign each utterance to an appropriate expert encoder. A frozen expert preserves high-resource language capabilities, while a trainable expert adapts to and specializes in medium- and low-resource languages. We further introduce a five-stage curriculum learning strategy that substantially reduces data dependence, requiring only 10 hours of paired S2TT data per language for effective alignment. We conduct extensive experiments on 45 languages, systematically evaluating all $45 \times 44$ translation directions. Our 4B-parameter model achieves state-of-the-art performance, outperforming substantially larger baselines. Empirical analyses show that MoSE improves high-, medium-, and low-resource languages simultaneously, with the largest gains on low-resource speech, thereby breaking the curse of multilinguality without compromising high-resource performance. To support future multilingual S2TT research, we release our code and models.

cs.CL↗

arXiv2026

Holomorphic functions on complete Hermitian manifolds with flat Chern connection

A classical result of Boothby states that any compact Hermitian manifold with flat Chern connection is covered by a complex Lie group. In this work, we prove a sharp generalization: the universal cover of any complete Hermitian manifold with vanishing Chern curvature and torsion of sublinear growth is holomorphically isometric to a complex Lie group equipped with a left-invariant metric. The proof relies on a new gradient estimate for holomorphic functions on complete Hermitian manifolds with nonnegative second Ricci curvature. Combining this estimate with methods from sub-Riemannian geometry, we establish quantitative characterizations of function theory on these manifolds.

math.DG↗

arXiv2026

Verifiable blind probabilistic error cancellation

Quantum error mitigation (QEM) is an essential tool for mitigating hardware noise without incurring space overhead. Yet, its reliability depends on modeling, calibration, and implementation, leaving end-to-end security on untrusted quantum hardware unresolved. We address this problem by introducing verifiable blind probabilistic error cancellation (VBPEC), the first secure verification protocol against a fully malicious adversary that integrates QEM. VBPEC brings probabilistic error cancellation (PEC), a widely studied QEM technique, within the scope of composable security by formalizing delegated mitigation as a cryptographic resource in the abstract cryptography framework. The protocol performs PEC with perfect blindness and an exponentially small security error. VBPEC retains the absence of quantum-space overhead from recent statistically-secure verified quantum computation protocols and from PEC. The only overhead takes the form of additional repetitions due to the QEM procedure. To achieve this, we extend trap-based verification from deterministic pass/fail checks to statistical tests that benefit from QEM and develop a new proof technique that integrates the corresponding additional deviation sources. Rather than merely tolerating honest noise below a fixed threshold, VBPEC actively cancels it, enabling correctly mitigated estimates to be accepted with high probability without compromising security. Our framework thus establishes an essential route towards secure, reliable, and practical delegated quantum computation on near-future quantum hardware: VBPEC fundamentally improves the practicality of verification.

quant-ph↗

arXiv2026

On Success and Simplicity: A Second Look at Transferable Vision-Language Attack Pipeline

Vision-Language Pre-training Models (VLPMs) are known to be vulnerable to adversarial attacks. Recent transferable attacks on VLPMs have followed a common pipeline with complicated loss functions or multi-stage text/image attacks. However, in this paper, we demonstrate that such a sophisticated attack pipeline can be simpler yet more successful. Specifically, we identify three previously overlooked issues caused by inappropriate cross-modal interactions and excessive operations. To address them, we propose the Simple Vision-Language Attack (SimVLA) pipeline, which observably improves transferability and efficiency. Experiments on four datasets and three downstream tasks validate the superiority of our pipeline. For instance, on Flickr30k text-image retrieval dataset, our SimVLA outperforms the SOTA baseline in R@1 transferability by 8.01\%-14.71\%, while consuming only about 35.73\% of the time and 46.26\% of the max VRAM. Overall, the superiority of our SimVLA highlights the importance of leveraging domain knowledge (e.g., our proposed cross-modal word identification), while blindly pursuing intricate operations (e.g, complex loss functions and redundant multi-stage designs) may even be harmful. We hope our SimVLA can serve as a simple yet effective backbone for future extensions. Code is available at https://github.com/RYC-98/SimVLA.

cs.CV↗

arXiv2026

Task-Oriented Sensing and Covert Transmissions for Collaborative Multi-AUV Systems

In underwater covert cooperative missions, autonomous underwater vehicles (AUVs) often cannot rely on active sonar to continuously obtain complete information, since active sensing and frequent communications increase the risk of exposure. As a result, AUVs primarily rely on passive observation, an approach that yields incomplete local perception and limited task efficiency. Although underwater acoustic communications can mitigate this limitation through information sharing, they are simultaneously constrained by long delays, severe interference, low reliability, and the risk of covert exposure. Existing communications-oriented multi-agent reinforcement learning (MARL) studies often model communication as an ideal information flow, whereas traditional communication optimization primarily focuses on link-level performance. However, both are insufficient to characterize the actual contribution of perceptual information to cooperative tasks under realistic conditions of covert physical communications. This paper proposes a Sensed Information Value Realization Multi-Agent Reinforcement Learning (SVR-MARL) framework that leverages practical information to characterize the utility of information for cooperative tasks and learns distributed cooperative policies under realistic communication and covert constraints. Through a case study of covert multi-AUV cooperative localization and tracking, the potential of the proposed framework to improve collaborative task efficiency while reducing unnecessary communication and exposure risks is demonstrated.

cs.LG↗

arXiv2026

Observational Evidence for Counter-helicity Magnetic Reconnection in a Solar Eruption

Magnetic reconnection between coronal magnetic systems carrying opposite self-helicity may play a role in solar eruptions, but observational evidence remains limited. We investigate an M7.0 flare in NOAA Active Region 13615 on 2024 March 28 using multiwavelength observations and nonlinear force-free field extrapolations. The reconstructed coronal field reveals a low-lying positive-helicity core field beneath an overlying magnetic system of opposite sign. During the eruption, the footpoint connectivity of these two magnetic systems changes markedly: field lines rooted in the western footpoint region change from positive to negative helicity, and the positive-helicity domain is substantially reduced. These changes are accompanied by a remote chromospheric brightening, intermittent EUV stripe-like brightenings extending from the source region toward the remote chromospheric brightening, the subsequent formation of large-scale coronal loops, and a weak outer hard X-ray source located at a footpoint of the core field. Together, these results suggest that the eruption was closely associated with reconnection between the core field and the overlying counter-helicity system, providing observational evidence that counter-helicity reconnection can contribute to the destabilization of eruptive solar magnetic fields.

astro-ph.SR↗

arXiv2026

Full-Pipeline Inference Optimization for MiMo-V2.5 Series: Pushing Hybrid SWA Efficiency to the Limit

We present a full-pipeline inference optimization for the MiMo-V2.5 model family, which combines Hybrid Sliding Window Attention (Hybrid SWA), sparse Mixture-of-Experts (MoE), and multimodal encoders. While Hybrid SWA can ideally reduce both attention compute and KVCache storage significantly compared to Full Attention, realizing these gains in production requires substantial engineering effort. We systematically optimize the KVCache system with layerwise prefetch, SWA-aware prefix cache trees, and specialized placement strategies, achieving strict $O(W)$ SWA storage and high cache hit rates. We further build GCache, a high-performance distributed cache infrastructure with RDMA-optimized networking, and develop a KVCache-affinity router to reduce computation while preserving load balancing. We also optimize for multimodal inputs, including GPU image preprocessing, parallel video decoding, and multimodal cache sharing. Together, these optimizations constitute the first large-scale LLM serving system in production that efficiently covers the Hybrid SWA + MoE + multimodal composite architecture.

cs.AR↗