Searcharxiv⌕ Search

arXiv subjects

Shuo Xu

Publications and source records attributed to Shuo Xu.

At least 19 recordsLinked to original sources
arXiv2026

Cluster finding with outskirt stellar masses and percolation

The abundance of galaxy clusters is a powerful cosmological probe, but optical cluster cosmology is limited by selection systematics, in particular the projection effects that affect cluster finders based on galaxy populations such as the red sequence. The outer stellar mass ($M_\mathrm{out}$) of cluster central galaxies -- e.g., the stellar mass in a 50-100 kpc annulus -- offers an alternative selection that relies only on the central galaxy and is therefore largely free from projection effects. Its primary systematic is instead satellite contamination, since massive clusters can host more than one galaxy with high outer stellar mass. Using the IllustrisTNG300 simulation at $z=0.4$, we quantify this contamination and investigate a simple, proximity-based percolation method to mitigate it, in which galaxies with lower outer stellar mass lying within a given radius of a more massive galaxy are removed from the sample. We find that the satellite fraction defined by the friends-of-friends (FoF) algorithm is modest even without percolation ($\leq 15\%$ for $M_\mathrm{out} > 10^{10}\,\mathrm{M}_\odot$ and $<10\%$ for $M_\mathrm{out} > 10^{11}\,\mathrm{M}_\odot$), and that percolation reduces it further, with the improvement increasing for percolation radii up to $3.0\,R_{200c}$. For a moderately high outer stellar mass cut ($\sim 4\times10^{10}\,\mathrm{M}_\odot$) and percolation radius ($\sim 2.0\,R_{200c}$), we recover a cluster sample that is both highly complete and pure for halo masses $\gtrsim 10^{14}\,\mathrm{M}_\odot$. These results indicate that outer stellar mass, combined with simple percolation, has the potential to provide a clean and readily calibratable selection of massive galaxy clusters.

astro-ph.CO↗
arXiv2026

SLO-Scaler: Uncertainty-Aware SLO-Driven Autoscaling for Microservices

Autoscaling microservice-based applications to satisfy Service Level Objectives (SLOs) remains challenging due to bursty workloads, cascading latency across service dependencies, and cold-start overhead. Existing approaches such as the Kubernetes Horizontal Pod Autoscaler (HPA) rely on threshold-based CPU or memory metrics, which react too slowly to traffic spikes. Recent predictive methods improve responsiveness but generate point forecasts that ignore prediction uncertainty, leading to over-provisioning or oscillatory scaling. We propose SLO-Scaler, an uncertainty-aware autoscaling framework that predicts short-horizon request rates, tail latency, and SLO violation probability using a Bayesian LSTM model. SLO-Scaler integrates confidence-interval-based scaling decisions with a dependency graph analysis module that localizes bottleneck services, avoiding unnecessary whole-chain scaling. We evaluate SLO-Scaler on the DeathStarBench Social Network benchmark deployed on Kubernetes under periodic, bursty, and long-tail traffic patterns. Under bursty traffic, SLO-Scaler reduces the SLO violation rate by 29-56%, lowers the average replica count by 18-33%, and decreases scaling event frequency by 38-59% compared with the baselines, while achieving lower tail latency.

cs.DC↗
arXiv2026

MagCeptor: Encoding Broadcast-Addressable Logic into Magnetic Receptors

Multicellular coordination relies on broadcast-addressable receptors, yet engineered magnetic systems face an addressability bottleneck because global fields intrinsically conflate power and control. Here, we introduce MagCeptors to resolve this by encoding selectivity directly into magnetic topology. Establishing an energetic isomorphism with biological receptors, these arrays utilize local couplings to shape potential landscapes where global field vectors act as spatial keys, triggering deterministic snap-through instabilities. This architecture decouples force from source distance, achieving a density of 385 mN/mm3 (>50-fold increase over prior art). We validate this primitive through signal demultiplexing, embodied sequential logic, and untethered distributed networking. This framework enables distributed systems to orchestrate complex tasks without tethers or electronics, relying solely on the intrinsic logic of matter.

eess.SY↗
arXiv2026

Rethinking Supervision Granularity: Segment-Level Learning for LLM-Based Theorem Proving

Automated theorem proving with large language models in Lean 4 is commonly approached through either step-level tactic prediction with tree search or whole-proof generation. These two paradigms represent opposite granularities for constructing supervised training data: the former provides dense local signals but may fragment coherent proof processes, while the latter preserves global structure but requires complex end-to-end generation. In this paper, we revisit supervision granularity as a training set construction problem over proof trajectories and propose segment-level supervision, a training data construction strategy that extracts locally coherent proof segments for training policy models. We further reuse the same strategy at inference time to trigger short rollouts for existing step-level models. When trained with segment-level supervision on STP, LeanWorkbook, and NuminaMath-LEAN, the resulting policy models achieve proof success rates of 64.84%, 60.90%, and 66.31% on miniF2F, respectively, consistently outperforming both step-level and whole-proof baselines. Goal-aware rollout further improves existing step-level provers while reducing inference costs. It increases the proof success rate of BFS-Prover-V2-7B from 68.77% to 70.74% and that of InternLM2.5-StepProver from 59.59% to 60.33%, showing that appropriate supervision granularity better aligns model learning with proof structure and search. Code and models are available at https://github.com/NJUDeepEngine/SEG-ATP.

cs.AI↗
arXiv2025

Reaching for the Edge II: Stellar Halos out to Large Radii as a Tracer of Dark Matter Halo Mass

The diffuse outskirts of brightest cluster galaxies (BCGs) encode valuable information about the assembly history and mass of their host dark matter halos. However, the low surface brightness of these stellar halos has historically made them difficult to observe. Recent deep imaging, particularly with Hyper Suprime-Cam (HSC), has shown that the stellar mass within relatively large projected annuli, such as within $50$ and $100$ kpc, is a promising proxy for halo mass. However, the optimal radial definition of this "outskirt mass" remains uncertain. We construct an HSC-like mock observing pipeline to measure the stellar mass density profiles of BCGs in the IllustrisTNG simulations. Our mock observations closely reproduce HSC profiles across six orders of magnitude in surface density. We then systematically measure stellar masses within different annuli and how tightly they are connected to halo mass. We find that stellar masses measured within simple apertures exhibit considerably more scatter in the stellar mass-halo mass relation than those measured within projected ellipsoidal annuli. We identify an optimal range of definitions, with inner radii between $\sim 70$-$200$ kpc and outer radii between $\sim 125$-$500$ kpc. We also introduce two halo-mass-dependent S\'ersic models for the average stellar halo profiles. We present a S\'ersic-based fitting function that describes the profiles as a function of the halo mass, $M_{\rm vir}$, with a median error of $54\%$. Adding the central stellar mass of the BCG as a second parameter slightly improves the accuracy to a median error of $39\%$. Together, these results provide fitting functions for BCG stellar halos that can be applied to future wide-field surveys to infer halo masses from deep imaging data.

astro-ph.GA↗
arXiv2025

Bridging Vision and Language: Optimal Transport-Driven Radiology Report Generation via LLMs

Radiology report generation represents a significant application within medical AI, and has achieved impressive results. Concurrently, large language models (LLMs) have demonstrated remarkable performance across various domains. However, empirical validation indicates that general LLMs tend to focus more on linguistic fluency rather than clinical effectiveness, and lack the ability to effectively capture the relationship between X-ray images and their corresponding texts, thus resulting in poor clinical practicability. To address these challenges, we propose Optimal Transport-Driven Radiology Report Generation (OTDRG), a novel framework that leverages Optimal Transport (OT) to align image features with disease labels extracted from reports, effectively bridging the cross-modal gap. The core component of OTDRG is Alignment \& Fine-Tuning, where OT utilizes results from the encoding of label features and image visual features to minimize cross-modal distances, then integrating image and text features for LLMs fine-tuning. Additionally, we design a novel disease prediction module to predict disease labels contained in X-ray images during validation and testing. Evaluated on the MIMIC-CXR and IU X-Ray datasets, OTDRG achieves state-of-the-art performance in both natural language generation (NLG) and clinical efficacy (CE) metrics, delivering reports that are not only linguistically coherent but also clinically accurate.

cs.CV↗
arXiv2025

Benchmarking Generalizable Bimanual Manipulation: RoboTwin Dual-Arm Collaboration Challenge at CVPR 2025 MEIS Workshop

Embodied Artificial Intelligence (Embodied AI) is an emerging frontier in robotics, driven by the need for autonomous systems that can perceive, reason, and act in complex physical environments. While single-arm systems have shown strong task performance, collaborative dual-arm systems are essential for handling more intricate tasks involving rigid, deformable, and tactile-sensitive objects. To advance this goal, we launched the RoboTwin Dual-Arm Collaboration Challenge at the 2nd MEIS Workshop, CVPR 2025. Built on the RoboTwin Simulation platform (1.0 and 2.0) and the AgileX COBOT-Magic Robot platform, the competition consisted of three stages: Simulation Round 1, Simulation Round 2, and a final Real-World Round. Participants totally tackled 17 dual-arm manipulation tasks, covering rigid, deformable, and tactile-based scenarios. The challenge attracted 64 global teams and over 400 participants, producing top-performing solutions like SEM and AnchorDP3 and generating valuable insights into generalizable bimanual policy learning. This report outlines the competition setup, task design, evaluation methodology, key findings and future direction, aiming to support future research on robust and generalizable bimanual manipulation policies. The Challenge Webpage is available at https://robotwin-benchmark.github.io/cvpr-2025-challenge/.

cs.RO↗
arXiv2025

TACOcc:Target-Adaptive Cross-Modal Fusion with Volume Rendering for 3D Semantic Occupancy

The performance of multi-modal 3D occupancy prediction is limited by ineffective fusion, mainly due to geometry-semantics mismatch from fixed fusion strategies and surface detail loss caused by sparse, noisy annotations. The mismatch stems from the heterogeneous scale and distribution of point cloud and image features, leading to biased matching under fixed neighborhood fusion. To address this, we propose a target-scale adaptive, bidirectional symmetric retrieval mechanism. It expands the neighborhood for large targets to enhance context awareness and shrinks it for small ones to improve efficiency and suppress noise, enabling accurate cross-modal feature alignment. This mechanism explicitly establishes spatial correspondences and improves fusion accuracy. For surface detail loss, sparse labels provide limited supervision, resulting in poor predictions for small objects. We introduce an improved volume rendering pipeline based on 3D Gaussian Splatting, which takes fused features as input to render images, applies photometric consistency supervision, and jointly optimizes 2D-3D consistency. This enhances surface detail reconstruction while suppressing noise propagation. In summary, we propose TACOcc, an adaptive multi-modal fusion framework for 3D semantic occupancy prediction, enhanced by volume rendering supervision. Experiments on the nuScenes and SemanticKITTI benchmarks validate its effectiveness.

cs.CV↗
arXiv2025

LLM-based Automated Theorem Proving Hinges on Scalable Synthetic Data Generation

Recent advancements in large language models (LLMs) have sparked considerable interest in automated theorem proving and a prominent line of research integrates stepwise LLM-based provers into tree search. In this paper, we introduce a novel proof-state exploration approach for training data synthesis, designed to produce diverse tactics across a wide range of intermediate proof states, thereby facilitating effective one-shot fine-tuning of LLM as the policy model. We also propose an adaptive beam size strategy, which effectively takes advantage of our data synthesis method and achieves a trade-off between exploration and exploitation during tree search. Evaluations on the MiniF2F and ProofNet benchmarks demonstrate that our method outperforms strong baselines under the stringent Pass@1 metric, attaining an average pass rate of $60.74\%$ on MiniF2F and $21.18\%$ on ProofNet. These results underscore the impact of large-scale synthetic data in advancing automated theorem proving.

cs.AI↗
arXiv2025

Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt Tuning

Modeling label correlations has always played a pivotal role in multi-label image classification (MLC), attracting significant attention from researchers. However, recent studies have overemphasized co-occurrence relationships among labels, which can lead to overfitting risk on this overemphasis, resulting in suboptimal models. To tackle this problem, we advocate for balancing correlative and discriminative relationships among labels to mitigate the risk of overfitting and enhance model performance. To this end, we propose the Multi-Label Visual Prompt Tuning framework, a novel and parameter-efficient method that groups classes into multiple class subsets according to label co-occurrence and mutual exclusivity relationships, and then models them respectively to balance the two relationships. In this work, since each group contains multiple classes, multiple prompt tokens are adopted within Vision Transformer (ViT) to capture the correlation or discriminative label relationship within each group, and effectively learn correlation or discriminative representations for class subsets. On the other hand, each group contains multiple group-aware visual representations that may correspond to multiple classes, and the mixture of experts (MoE) model can cleverly assign them from the group-aware to the label-aware, adaptively obtaining label-aware representation, which is more conducive to classification. Experiments on multiple benchmark datasets show that our proposed approach achieves competitive results and outperforms SOTA methods on multiple pre-trained models.

cs.CV↗
arXiv2025

Relationship between 2D and 3D Galaxy Stellar Mass and Correlations with Halo Mass

Recent studies suggest that the stars in the outer regions of massive galaxies trace halo mass better than the inner regions and that an annular stellar mass provides a low scatter method of selecting galaxy clusters. However, we can only observe galaxies as projected two-dimensional objects on the sky. In this paper, we use a sample of simulated galaxies to study how well galaxy stellar mass profiles in three dimensions correlate with halo mass, and what effects arise when observationally projecting stellar profiles into two dimensions. We compare 2D and 3D outer stellar mass selections and find that they have similar performance as halo mass proxies and that, surprisingly, a 2D selection sometimes has marginally better performance. We also investigate whether the weak lensing profiles around galaxies selected by 2D outer stellar mass suffer from projection effects. We find that the lensing profiles of samples selected by 2D and 3D definitions are nearly identical, suggesting that the 2D selection does not create a bias. These findings underscore the promise of using outer stellar mass as a tool for identifying galaxy clusters.

astro-ph.CO↗
arXiv2024

The Outskirt Stellar Mass of Low-Redshift Massive Galaxies is an Excellent Halo Mass Proxy in Illustris/IllustrisTNG Simulations

Recent observations suggest that the extended stellar halos of low-redshift massive galaxies are tightly connected to the assembly of their dark matter halos. In this paper, we use the Illustris, IllustrisTNG100, and IllustrisTNG300 simulations to compare how different stellar aperture masses trace halo mass. For massive central galaxies ($M_\star\geq 10^{11.2}M_\odot$), we find that a 2D outskirt stellar mass measured between 50 to 100 kpc ($M_{\star,[50,100]}$) consistently outperforms other aperture-based stellar masses. We further show that $M_{\star,[50,100]}$ correlates better with halo mass than the total amount of accreted stars (the ex situ mass), which suggests that not all accreted stars connect to halo assembly equally. While the galaxy formation recipes are different between Illustris and IllustrisTNG100, the two simulations yield consistent ex situ outskirt fractions for massive galaxies (about 70% in $M_{\star,[50,100]}$). These results demonstrate the potential of using the outskirt stellar mass to deepen our understanding of galaxy-halo connection in massive dark matter halos and trace dark matter halos better.

astro-ph.GA↗
arXiv2024

Headache to Overstock? Promoting Long-tail Items through Debiased Product Bundling

Product bundling aims to organize a set of thematically related items into a combined bundle for shipment facilitation and item promotion. To increase the exposure of fresh or overstocked products, sellers typically bundle these items with popular products for inventory clearance. This specific task can be formulated as a long-tail product bundling scenario, which leverages the user-item interactions to define the popularity of each item. The inherent popularity bias in the pre-extracted user feedback features and the insufficient utilization of other popularity-independent knowledge may force the conventional bundling methods to find more popular items, thereby struggling with this long-tail bundling scenario. Through intuitive and empirical analysis, we navigate the core solution for this challenge, which is maximally mining the popularity-free features and effectively incorporating them into the bundling process. To achieve this, we propose a Distilled Modality-Oriented Knowledge Transfer framework (DieT) to effectively counter the popularity bias misintroduced by the user feedback features and adhere to the original intent behind the real-world bundling behaviors. Specifically, DieT first proposes the Popularity-free Collaborative Distribution Modeling module (PCD) to capture the popularity-independent information from the bundle-item view, which is proven most effective in the long-tail bundling scenario to enable the directional information transfer. With the tailored Unbiased Bundle-aware Knowledge Transferring module (UBT), DieT can highlight the significance of popularity-free features while mitigating the negative effects of user feedback features in the long-tail scenario via the knowledge distillation paradigm. Extensive experiments on two real-world datasets demonstrate the superiority of DieT over a list of SOTA methods in the long-tail bundling scenario.

cs.IR↗
arXiv2024

MAC: A Benchmark for Multiple Attributes Compositional Zero-Shot Learning

Compositional Zero-Shot Learning (CZSL) aims to learn semantic primitives (attributes and objects) from seen compositions and recognize unseen attribute-object compositions. Existing CZSL datasets focus on single attributes, neglecting the fact that objects naturally exhibit multiple interrelated attributes. Their narrow attribute scope and single attribute labeling introduce annotation biases, misleading the learning of attributes and causing inaccurate evaluation. To address these issues, we introduce the Multi-Attribute Composition (MAC) dataset, encompassing 22,838 images and 17,627 compositions with comprehensive and representative attribute annotations. MAC shows complex relationship between attributes and objects, with each attribute type linked to an average of 82.2 object types, and each object type associated with 31.4 attribute types. Based on MAC, we propose multi-attribute compositional zero-shot learning that requires deeper semantic understanding and advanced attribute associations, establishing a more realistic and challenging benchmark for CZSL. We also propose Multi-attribute Visual-Primitive Integrator (MVP-Integrator), a robust baseline for multi-attribute CZSL, which disentangles semantic primitives and performs effective visual-primitive association. Experimental results demonstrate that MVP-Integrator significantly outperforms existing CZSL methods on MAC with improved inference efficiency.

cs.CV↗
arXiv2024

High Spectral-Efficiency, Ultra-low MIMO SDM Transmission over a Field-Deployed Multi-Core OAM Fiber

Few-mode multi-core fiber (FM-MCF) based Space-Division Multiplexing (SDM) systems possess the potential to maximize the number of multiplexed spatial channels per fiber by harnessing both the space (fiber cores) and mode (optical mode per core) dimensions. However, to date, no SDM transmissions over field-deployed FM-MCFs in realistic outdoor settings have been reported, which contrasts with SDM schemes demonstrated using single-mode multi-core fibers (SM-MCFs) installed in practical fiber cable ducts. In this paper, we present the successful demonstration of bidirectional SDM transmission over a 5-km field-deployed seven ring-core fiber (7-RCF) with a cladding diameter of 178 ${\mu}$m, achieving a Spectral Efficiency (SE) of 2$\times$201.6 bit/s/Hz. This work establishes a new record for the highest SE attained in SDM demonstrations utilizing field-deployed fiber cables, achieving an approximate 10x increase compared to the SE of reported field-deployed optical fiber cable transmission systems. Notably, these results are realized through the utilization of small-scale modular 4$\times$4 multiple-input multiple-output (MIMO) processing with a time-domain equalization (TDE) tap number not exceeding 15, maintaining a complexity per unit capacity comparable to that of MIMO equalization in SDM demonstrations employing weakly coupled SM-MCF cables. These results underscore the significant potential for achieving heightened SE and expanding capacity per individual fiber using SDM techniques in practical applications.

cs.NI↗
arXiv2024

SimFair: Physics-Guided Fairness-Aware Learning with Simulation Models

Fairness-awareness has emerged as an essential building block for the responsible use of artificial intelligence in real applications. In many cases, inequity in performance is due to the change in distribution over different regions. While techniques have been developed to improve the transferability of fairness, a solution to the problem is not always feasible with no samples from the new regions, which is a bottleneck for pure data-driven attempts. Fortunately, physics-based mechanistic models have been studied for many problems with major social impacts. We propose SimFair, a physics-guided fairness-aware learning framework, which bridges the data limitation by integrating physical-rule-based simulation and inverse modeling into the training design. Using temperature prediction as an example, we demonstrate the effectiveness of the proposed SimFair in fairness preservation.

cs.LG↗
arXiv2023

Deep Radon Prior: A Fully Unsupervised Framework for Sparse-View CT Reconstruction

Although sparse-view computed tomography (CT) has significantly reduced radiation dose, it also introduces severe artifacts which degrade the image quality. In recent years, deep learning-based methods for inverse problems have made remarkable progress and have become increasingly popular in CT reconstruction. However, most of these methods suffer several limitations: dependence on high-quality training data, weak interpretability, etc. In this study, we propose a fully unsupervised framework called Deep Radon Prior (DRP), inspired by Deep Image Prior (DIP), to address the aforementioned limitations. DRP introduces a neural network as an implicit prior into the iterative method, thereby realizing cross-domain gradient feedback. During the reconstruction process, the neural network is progressively optimized in multiple stages to narrow the solution space in radon domain for the under-constrained imaging protocol, and the convergence of the proposed method has been discussed in this work. Compared with the popular pre-trained method, the proposed framework requires no dataset and exhibits superior interpretability and generalization ability. The experimental results demonstrate that the proposed method can generate detailed images while effectively suppressing image artifacts.Meanwhile, DRP achieves comparable or better performance than the supervised methods.

eess.IV↗