arXiv subjects

Xiaofei Wang

Publications and source records attributed to Xiaofei Wang.

At least 19 recordsLinked to original sources

arXiv2026

Power and Sample Size Calculations for Hybrid Controlled Trials

Hybrid controlled trials (HCTs) augment randomized controls with external controls (ECs) to address practical challenges in randomized controlled trials (RCTs) and improve statistical power in settings such as rare diseases, oncology, and pediatrics. However, prospective sample-size determination is challenging because the required RCT sample size depends on the comparability of ECs with RCT controls, which are unavailable at the planning stage. We propose a 5+3 design for HCT sample-size determination based on an inverse probability weighting estimator of the average treatment effect. The framework uses five conventional RCT design parameters and three additional scalar parameters characterizing EC comparability: the number of outcome-drift-free ECs, an overlap coefficient for the covariate distributions of the RCT and ECs, and a correlation coefficient linking the sampling mechanism to the control potential outcome. We establish the asymptotic distribution of the estimator and prove that its variance is determined by these design parameters under the proposed working models, yielding sample-size calculations for both continuous and binary outcomes. Simulation studies evaluate finite-sample performance, and a real clinical application illustrates its practical use. The method is implemented in the hctdesign R package.

stat.ME↗

arXiv2026

Rubrics as Visual-Repair Context for Self-Evolving UI-to-Code Generation

Large vision-language models have shown strong progress in UI-to-code generation, yet their test-time self-evolution remains unstable. We first identify a fundamental obstacle, termed visual repair coupling: a local code edit may propagate through layout, style, and component dependencies, correcting one visual mismatch while degrading regions that were previously faithful. To address this issue, we present RubSE, a Rubric-guided Self-Evolution framework that uses rubrics to represent visual feedback as a structured visual-repair context. At each refinement round, RubSE generates typed candidate rubrics, selects one prioritized repair target, and stores previously selected rubrics as history, thereby steering each revision toward a well-scoped visual repair while discouraging repeated or over-broad changes. Evaluations across six VLMs and three UI-to-code benchmarks demonstrate that RubSE substantially outperforms na\"ive self-evolution in final-round and best-round settings, achieving more stable refinement trajectories and a higher trajectory-level performance ceiling. Further analysis shows that RubSE mitigates trajectory collapse by improving recovery from severe visual regressions, and that stronger rubric generators can transfer effective visual-repair guidance to weaker code improvers.

cs.CV↗

arXiv2026

Myelin Distribution at the Optic Nerve Myelination Transition Zone Influences Axonal Biomechanics

Purpose: The lamina cribrosa (LC) is considered the initial site of glaucomatous retinal ganglion cell (RGC) injury, and is also the region where unmyelinated RGC axons become myelinated. Here we sought to use finite element (FE) modeling to investigate how the configuration of the myelination transition zone (MTZ) influences the mechanical insult to RGC axons. Methods: A multiscale FE framework was developed to investigate the biomechanical effect of myelin distribution on IOP-induced axonal stress and strain at the MTZ. An anatomically based macro-scale FE eye model was used to compute LC deformations under 15 and 45 mmHg IOP. These deformations were then applied to micro-scale models of the posterior LC, consisting of axons, myelin sheaths, and surrounding matrix. Four distinct MTZ boundary configurations were simulated: one flat and three with random posterior offsets of 3, 6, or 9 {\mu}m, representing potential physiological variations. IOP-induced effective axonal strains and stresses were quantified across the different MTZ configurations. Results: Under IOP loading, axons exhibited longitudinal compression and transverse stretch, with marked effective stress and strain discontinuities at the myelin boundary. Across all models, the unmyelinated region exhibited higher effective stress and strain than the myelinated region, and this mechanical discontinuity increased with larger MTZ offsets. Conclusions: Glaucoma-associated demyelination has been previously suggested to precede RNFL thinning. Here we have shown that the MTZ configuration directly influences RGC axonal mechanics. Whether different MTZ profiles can initiate glaucomatous injury, whether demyelination accelerates disease progression, or whether both mechanisms contribute, remains to be determined.

cs.CE↗

arXiv2026

From Network Automation to Trustworthy Autonomous Networking in the LLM Era: A Network Control Intelligence Perspective

Since the inception of modern communication networks, the quest for operations automation has never ceased. Yet the evolution of network automation is difficult to characterize with a single maturity ladder. Throughout this history, network control systems have expanded their capabilities for observation, decision support, routine execution, and operator interaction, but these capabilities have not advanced uniformly. Such uneven progress makes the degree of automation an unreliable proxy for trustworthy network-side actuation. The unresolved question is not simply how much automation a system provides, but under what conditions it can be entrusted to change the network state. This paper examines that question through Network Control Intelligence (NCI), a five-axis framework spanning Decision Logic, Adaptability, Knowledge, Control Delegation, and Interface. We use NCI to organize the evolution of network-control systems into three eras: rule-based and scripted automation, programmable and data-driven control, and Large Language Model (LLM)-enabled network operations. Viewed through this framework, the three eras reveal a persistent asymmetry. None of these gains, however, automatically determines when network control should be trusted to change the network state. We frame trustworthy autonomy as a governed alignment between what a system can infer, what it can verify, and what it is authorized to execute. On that basis, the paper develops a reference architecture that separates proposal generation from governed execution, identifies recurring integration patterns for LLM-enabled operations, and derives a research agenda for higher network autonomy under explicit assurance, safety, and governance constraints.

cs.NI↗

arXiv2026

Robust Estimation and Inference with Selective Borrowing in Hybrid Controlled Trials: A Tutorial with SelectiveIntegrative and intFRT

Hybrid controlled trials (HCTs) augment randomized controlled trials (RCTs) with external controls (ECs) to improve statistical efficiency when RCTs face limited sample sizes, slow accrual, or ethical constraints. However, valid use of ECs requires careful adjustment for covariate shift and outcome drift, as inappropriate borrowing may introduce bias and compromise inference. This tutorial provides a practical workflow for estimation and inference in HCTs. We first present a statistical analysis roadmap covering estimands, identification assumptions, eligibility alignment, matching, full and selective borrowing strategies, and both asymptotic inference and randomization tests. We then demonstrate step-by-step implementation using the SelectiveIntegrative and intFRT packages. The workflow is illustrated using a synthetic lung cancer dataset included in the intFRT package that mimics the CALGB 9633 trial and ECs from the National Cancer Database. The tutorial aims to help applied statisticians conduct transparent, interpretable, and reproducible HCT analyses that improve efficiency while maintaining valid inference.

stat.ME↗

arXiv2026

A Conformal Selection Framework for Individual Treatment Beneficiaries with Auxiliary External Data

Identifying patients who are likely to benefit from a treatment is central to precision medicine and can guide follow-up trials, enrichment designs, and individualized decisions. Although randomized controlled trials (RCTs) provide evidence on efficacy, they are usually powered to estimate average treatment effects rather than patient-level benefit. Meanwhile, artificial intelligence and machine learning methods offer flexible tools for estimating heterogeneous treatment effects, especially when augmented by real-world data (RWD). However, in practice, these estimated effects are often translated into decisions through simple ranking or thresholding rules, which can ignore uncertainty and multiplicity when many patients are evaluated simultaneously. Motivated by this, we propose a model-agnostic conformal inference framework for uncertainty-aware beneficiary selection. The framework reformulates CATE-based treatment-benefit selection as a multiple-testing problem. For each candidate, we test whether the conditional treatment benefit exceeds a clinically meaningful threshold and construct a conformal p-value using RCT-based calibration. These p-values are then adjusted by the Benjamini-Hochberg procedure to control the false discovery rate (FDR) among selected beneficiaries. To improve efficiency when RCT sample sizes are limited, external data, such as RWD, can be used to train flexible treatment effect models, while conformal calibration remains anchored in the RCT data. It can be paired with conventional machine learning algorithms and emerging tabular foundation models. Simulations show that the framework maintains FDR control, with power depending on the base learner and external-data comparability. A case study in early-stage non-small-cell lung cancer illustrates how the method identifies candidate profiles with evidence of benefit from limited resection to reduce overtreatment.

stat.ME↗

arXiv2026

Preserving Speech-to-Text LLM Capabilities in Speech-to-Speech Generation

Strong speech-to-text (S2T) LLMs already provide robust speech perception and text reasoning, but adding speech-to-speech (S2S) output is challenging: fine-tuning the backbone can degrade the original S2T performance, while attaching a downstream talker reintroduces a serial text-to-speech bottleneck. We present PRIME-Speech, a frozen-backbone S2S conversion framework that trains only speech-generation modules. PRIME-Speech synchronizes a causal audio post-decoder with intermediate hidden states of the frozen backbone, so codec tokens are generated from the model's evolving reasoning trajectory rather than from completed text chunks. The post-decoder uses mixed hidden-state, text, and audio-history conditioning, and a training-time packing strategy with turn-level audio KV-cache and position reset stabilizes multi-turn spoken interaction without additional multi-turn S2S training data. Multi-token prediction further reduces the effective codec prediction rate and improves first-audio latency without modifying the reasoning path. Across speech translation, spoken QA, speech understanding, and multi-turn dialogue, PRIME-Speech preserves the S2T behavior of the frozen backbone while producing accurate, low-WER spoken responses.

eess.AS↗

arXiv2026

LLM can Read Spectrogram: Encoder-free Speech-Language Modeling

Recent speech-aware large language models (Speech-LLMs) rely on pre-trained speech encoders to convert audio into semantic/acoustic rich representations consumable by LLM. In this work, instead, we explore: can an LLM learn to read Mel spectrogram directly without a dedicated speech encoder? We propose Mel-LLM, an encoder-free Speech-LLM that feeds lightly pre-processed Mel-spectrogram patches directly into the LLM through a linear projection, allowing the LLM to learn speech-text alignment purely through its own parameters. We focus on speech understanding tasks, including automatic speech recognition (ASR), spoken QA and audio understanding. For ASR, we evaluate on the OpenASR Leaderboard public sets and production-level scaling experiments, demonstrating that the encoder-free solution achieves competitive performance with only limited degradation compared to encoder-initialized counterparts. We find that when data is limited, initialization from a multimodal checkpoint (Phi-4-MM) is crucial for maintaining performance. We also present ablation studies suggesting which LLM layers are most involved in speech adaptation. Beyond ASR, we extend Mel-LLM with general speech/audio understanding tasks, revealing an acoustic-semantic trade-off: directly exposing the LLM to Mel-spectrogram input improves paralinguistic and non-ASR acoustic tasks, while knowledge-intensive spoken QA remains more challenging than encoder-anchored systems. We additionally include a text-to-speech (TTS) proof-of-concept with a next-token VAE decoder, showing that direct Mel generation is possible but still trails stronger latent-diffusion generation.

eess.AS↗

arXiv2026

Partially Observable Adversarial Patch Attacks on Vision-Language-Action Models in Robotics

Vision-language-action (VLA) models are gaining attention in robotics, yet their robustness to adversarial attacks remains largely unexplored. Existing work shows that adversarial patches can mislead VLA-based robots but assumes full access to the entire execution trajectory, an unrealistic requirement in practice. We address this limitation by formulating a partially observable threat model, where the adversary can exploit only a short prefix of the trajectory to generate a fixed patch applied to all subsequent frames. Under this setting, we propose a two-phase framework. First, we localize the patch using the model's attention maps to identify visually critical regions that correspond to the full instruction. Then, we optimize the patch to disrupt the semantic grounding of target objects and increase the curvature of action trajectories, thereby compounding failures in both perception and control. Extensive experiments in simulation and real-world robotic environments show that our method sustains adversarial effects under partial observability, inducing long-horizon disruptions and significantly reducing task success rates.

cs.RO↗

arXiv2026

StateXDiff: Cell State-Contextualized Multimodal Diffusion for Single-Cell Perturbation Prediction

Predicting drug-induced cellular state changes at single-cell resolution remains a central challenge in virtual cell modeling, particularly under out-of-distribution (OOD) conditions. Current approaches predominantly rely on RNA-based assays, which often fail to adequately capture the diverse cellular states underlying drug responses. Moreover, conditional distribution shifts and low signal-to-noise ratios frequently cause models to learn spurious correlations rather than genuine state transitions. To address these limitations, we introduce StateXDiff, a cell State-contextualized multimodal (X) Diffusion framework for predicting single-cell responses to drug perturbations. The framework operates sequentially: first, it learns a disentangled, multimodal representation of cellular state by integrating transcriptomic profiles with inferred protein features; second, it employs a conditional diffusion model to generate perturbation-specific changes. Our approach introduces a Virtual Multimodal Cell State, which augments RNA-based representations with protein-level context, and a Mechanism-aware Drug-Gene Template, which consolidates multi-source biological knowledge for accurate drug representation. Generation is driven by a latent-space diffusion Transformer, regularized through quality-aware triplet constraints, including positive drug-protein pairs or protein-drug mismatched pairs, and explicit protein-reliability weighting. Extensive evaluation demonstrates that StateXDiff consistently enhances generalization performance across three challenging settings: unseen cell lines, unseen drugs, and combinatorial perturbations.

q-bio.GN↗

arXiv2026

FineServe: A Fine-Grained Dataset and Characterization of Global LLM Serving Workloads

Large language models (LLMs) are increasingly deployed as always-on online services, making efficient LLM serving a critical systems challenge. Achieving low latency and high throughput under volatile demand requires deep understanding of real-world serving workloads, yet existing studies often rely on proxy traces or coarse-grained characterizations that fail to capture the heterogeneity of modern multi-model LLM platforms. We present FineServe, an in-the-wild, multi-model LLM serving workload dataset collected from a global commercial marketplace, enabling fine-grained characterization of real-world serving dynamics across heterogeneous models and tasks. Leveraging FineServe, we conduct a comprehensive analysis of arrival dynamics and token behavior, revealing fundamentally different fluctuation regimes across model architectures, scales and task intents. Building on these insights, we develop the FineServe workload generator, which composes fine-grained model-aware workloads into configurable mixtures tailored for benchmarking multi-model serving platforms. By exposing these fine-grained workload dynamics, FineServe provides a realistic foundation for evaluating routing, scheduling, and capacity-planning strategies in LLM serving systems. FineServe is available at https://github.com/hihiztc1/FineServe.

cs.AI↗

arXiv2026

CoLLM: Continuous Adaptation for SLO-Aware LLM Serving on Shared GPU Clusters

As Large Language Models (LLMs) are increasingly adopted in edge intelligence to power domain-specific applications and personalized services, the quality and efficiency of the LLM post-training phase-including fine-tuning and inference, have become critical due to constrained resources. Although recent advances in federated parameter-efficient fine-tuning (FL PEFT) and low-latency inference have improved individual task performance, fine-tuning and inference are still handled as isolated workloads, which overlooks their interdependence and results in redundant deployments and delayed improvement in inference quality. To address these limitations, we introduce a new co-execution framework and instantiate it with CoLLM, a system that unifies FL PEFT and inference on shared edge replicas and model parameters. CoLLM addresses key challenges at both replica and cluster levels through: (1) an intra-replica model sharing mechanism that enables real-time model parameter reuse via unmerged inference and shadow adapter strategies; and (2) a two-timescale inter-replica coordination algorithm that adaptively balances fine-tuning and inference workloads to jointly optimize long-term model quality gains and short-term inference efficiency. Extensive evaluation across diverse LLMs and real-world traces show that CoLLM consistently outperforms state-of-the-art LLM systems, achieving up to 3x higher goodput, demonstrating its effectiveness in enabling seamless LLM post-training for edge intelligence.

cs.DC↗

arXiv2026

Selective Information Borrowing for Region-Specific Treatment Effect Inference under Covariate Mismatch in Multi-Regional Clinical Trials

Multi-regional clinical trials (MRCTs) are central to global drug development, enabling evaluation of treatment effects across diverse populations. A key challenge is valid and efficient inference for a region-specific estimand when the target region is small and differs from auxiliary regions in baseline covariates or unmeasured factors. We adopt an estimand-based framework and focus on the region-specific average treatment effect (RSATE) in a prespecified target region, which is directly relevant to local regulatory decision-making. Cross-region differences can induce covariate shift, covariate mismatch, and outcome drift, potentially biasing information borrowing and invalidating RSATE inference. To address these issues, we develop a unified causal inference framework with selective information borrowing. First, we introduce an inverse-variance weighting estimator that combines a "small-sample, rich-covariate" target-only estimator with a "large-sample, limited-covariate" full-borrowing doubly robust estimator, maximizing efficiency under no outcome drift. Second, to accommodate outcome drift, we apply conformal prediction to assess patient-level comparability and adaptively select auxiliary-region patients for borrowing. Third, to ensure rigorous finite-sample inference, we employ a conditional randomization test with exact, model-free, selection-aware type I error control. Simulation studies show the proposed estimator improves efficiency, yielding 10-50% reductions in mean squared error and higher power relative to no-borrowing and full-borrowing approaches, while maintaining valid inference across diverse scenarios. An application to the POWER trial further demonstrates improved precision for RSATE estimation.

stat.ME↗

arXiv2026

Rethinking Transferable Adversarial Attacks on Point Clouds from a Compact Subspace Perspective

Transferable adversarial attacks on point clouds remain challenging, as existing methods often rely on model-specific gradients or heuristics that limit generalization to unseen architectures. In this paper, we rethink adversarial transferability from a compact subspace perspective and propose CoSA, a transferable attack framework that operates within a shared low-dimensional semantic space. Specifically, each point cloud is represented as a compact combination of class-specific prototypes that capture shared semantic structure, while adversarial perturbations are optimized within a low-rank subspace to induce coherent and architecture-agnostic variations. This design suppresses model-dependent noise and constrains perturbations to semantically meaningful directions, thereby improving cross-model transferability without relying on surrogate-specific artifacts. Extensive experiments on multiple datasets and network architectures demonstrate that CoSA consistently outperforms state-of-the-art transferable attacks, while maintaining competitive imperceptibility and robustness under common defense strategies. Codes will be made public upon paper acceptance.

cs.CV↗

arXiv2026

Optimal Transport-Induced Samples against Out-of-Distribution Overconfidence

Deep neural networks (DNNs) often produce overconfident predictions on out-of-distribution (OOD) inputs, undermining their reliability in open-world environments. Singularities in semi-discrete optimal transport (OT) mark regions of semantic ambiguity, where classifiers are particularly prone to unwarranted high-confidence predictions. Motivated by this observation, we propose a principled framework to mitigate OOD overconfidence by leveraging the geometry of OT-induced singular boundaries. Specifically, we formulate an OT problem between a continuous base distribution and the latent embeddings of training data, and identify the resulting singular boundaries. By sampling near these boundaries, we construct a class of OOD inputs, termed optimal transport-induced OOD samples (OTIS), which are geometrically grounded and inherently semantically ambiguous. During training, a confidence suppression loss is applied to OTIS to guide the model toward more calibrated predictions in structurally uncertain regions. Extensive experiments show that our method significantly alleviates OOD overconfidence and outperforms state-of-the-art methods.

cs.CV↗

arXiv2025

Less Is More: Sparse and Cooperative Perturbation for Point Cloud Attacks

Most adversarial attacks on point clouds perturb a large number of points, causing widespread geometric changes and limiting applicability in real-world scenarios. While recent works explore sparse attacks by modifying only a few points, such approaches often struggle to maintain effectiveness due to the limited influence of individual perturbations. In this paper, we propose SCP, a sparse and cooperative perturbation framework that selects and leverages a compact subset of points whose joint perturbations produce amplified adversarial effects. Specifically, SCP identifies the subset where the misclassification loss is locally convex with respect to their joint perturbations, determined by checking the positivedefiniteness of the corresponding Hessian block. The selected subset is then optimized to generate high-impact adversarial examples with minimal modifications. Extensive experiments show that SCP achieves 100% attack success rates, surpassing state-of-the-art sparse attacks, and delivers superior imperceptibility to dense attacks with far fewer modifications.

cs.CR↗

arXiv2025

TechImage-Bench: Rubric-Based Evaluation for Technical Image Generation

We study technical image generation, where a model must synthesize information-dense, scientifically precise illustrations from detailed descriptions rather than merely produce visually plausible pictures. To quantify the progress, we introduce TechImage-Bench, a rubric-based benchmark that targets biology schematics, engineering/patent drawings, and general technical illustrations. For 654 figures collected from real textbooks and technical reports, we construct detailed image instructions and a hierarchy of rubrics that decompose correctness into 6,076 criteria and 44,131 binary checks. Rubrics are derived from surrounding text and reference figures using large multimodal models, and are evaluated by an automated LMM-based judge with a principled penalty scheme that aggregates sub-question outcomes into interpretable criterion scores. We benchmark several representative text-to-image models on TechImage-Bench and find that, despite strong open-domain performance, the best base model reaches only 0.801 rubric accuracy and 0.576 criterion score overall, revealing substantial gaps in fine-grained scientific fidelity. Finally, we show that the same rubrics provide actionable supervision: feeding failed checks back into an editing model for iterative refinement boosts a strong generator from 0.660 to 0.865 in rubric accuracy and from 0.382 to 0.697 in criterion score. TechImage-Bench thus offers both a rigorous diagnostic for technical image generation and a scalable signal for improving specification-faithful scientific illustrations.

cs.CV↗

arXiv2025

An Estimand-Focused Approach for AUC Generalization and Cross-Study Benchmarking

The area under the ROC curve (AUC) is the standard measure of a biomarker's discriminatory accuracy; however, AUC is rarely treated as a population-specific estimand. When validation cohorts differ from the intended target population in case mix, Na\"ive AUC estimates can mislead both generalization and cross-study comparison. We develop an estimand-focused framework that anchors biomarker AUC inference to a prespecified target population, aligning with the ICH E9(R1) estimand perspective adapted to discrimination rather than treatment effect. The framework supports two scientific goals: generalizing a study-specific AUC to a clinically relevant target population, and benchmarking AUCs across studies on a common population footing. Methodologically, we extend calibration weighting to the U-statistic formulation of AUC, allowing valid estimation even when the target population is characterized only by summary-level covariate information. This setting is common in biomarker validation, where individual-level target data are often unavailable and existing transportability methods may not be applicable. When patient-level real-world data are accessible, the proposed augmented variants provide double robustness and improved efficiency. We establish asymptotic properties and study their performances through comprehensive simulations. Furthermore, we demonstrate the proposed framework on the POWER trials, evaluating baseline stair-climb power (SCP) as a prognostic marker for 6-month survival in advanced non-small-cell lung cancer (NSCLC). Unlike prior work on transporting model-based predictive accuracy, our framework targets the biomarker-level estimand directly and addresses cross-study comparability - an issue not resolved by current methods.

stat.ME↗