Searcharxiv⌕ Search

arXiv subjects

Xuan Wang

Publications and source records attributed to Xuan Wang.

At least 19 recordsLinked to original sources
arXiv2026

Spatial Organization Is Accessibility Structure: Hierarchical Screen Reading of Mobile UIs

Screen readers for blind and low-vision (BLV) users flatten mobile UIs into a linear sequence. Modern mobile UIs, however, contain dozens of elements grouped by spatial layout. Such linear access is ill-suited for BLV users, who must infer each element's meaning and which elements form a functional unit. Drawing on a quantitative TalkBack audit and formative interviews, we present Spatial Screen Reader (SeeR), which builds a function-aware hierarchy from accessibility metadata and provides unit summaries with hierarchical navigation. Compared with TalkBack, SeeR reduces non-informative focus stops by 82.1% and navigation steps by 52.7%. With ten BLV participants, summaries helped them connect controls to their functions and skip nonessential elements, especially on unfamiliar pages, though extra navigation levels added effort when the target was already known. These findings recast spatial organization as accessibility information: apps and platforms should expose it, rather than leaving screen readers to infer it post hoc.

cs.HC↗
arXiv2026

Elastic Spectral State Space Models for Train-Once Budgeted Inference

Modern sequence models are typically trained at a fixed computational capacity, while real-world applications require deployment across platforms with different resource constraints. Existing approaches either train or distill separate compact models for selected budgets, or use elastic architectures that shrink dimensions such as width, depth, Feed-Forward Network (FFN) size, or SSM state dimension. In this paper, we propose Elastic Spectral State Space Models (ES-SSM), a train-once, export-many sequence modeling framework that gains elasticity through spectral approximation of the SSM sequence operator. ES-SSM builds on Hankel spectral filtering for state space models, where long-range token mixing is represented through fixed Hankel spectral channels that define an operator-level approximation resolution. To make this spectral elasticity reliably deployable, ES-SSM combines input-adaptive channel-wise gates with budget dropout, training the same spectral prefixes that are used by compact models at inference time. This encourages low-index channels to become predictive on their own, while higher-index channels provide refinements for larger budgets. We evaluate ES-SSM across byte-level language modeling, Long Range Arena, Speech Commands V2, and offline reinforcement learning benchmarks. Across these settings, a single trained ES-SSM can be truncated to competitive compact models compared with modern Transformer and SSM baselines at similar parameter scales. Furthermore, by testing under various runtime budgets, we observe smooth and stable quality-cost curves over a wide range of truncation levels. Code is available at https://github.com/GMU-MICO/ES-SSM.

cs.LG↗
arXiv2026

Sandman: A Sandbox Service for Production Agentic Reinforcement Learning

Agentic reinforcement learning (RL) needs stateful environments for tool execution and evaluation. Training jobs request thousands of environments in bursts that can overwhelm shared cluster services, while cancelled or timed-out requests can leave resources running. We present Sandman, a sandbox service built and operated at Amazon for large-scale agentic RL workloads in production on shared Kubernetes. A small daemon mounted into unmodified task images preserves files and processes across calls. Sandman paces creation requests, checks sandbox readiness, and retains cleanup responsibility after callers time out or cancel. Production deployments created 3.1 million sandboxes for 122 coding-agent training runs in two weeks. Peak production concurrency was 11,858 sandboxes. In one 10,000-request multi-client burst, Sandman made half the sandboxes usable in 116 seconds and all in 308 seconds. Full delivery took 44% less time than a tuned configuration of agent-sandbox, an open-source sandbox manager for Kubernetes. Sandman left no sandboxes at client exit. On software-engineering task images, median no-op command latency was 2.0 ms, compared with 14.3 ms for agent-sandbox and 203.5 ms for OpenSandbox. We also measure provisioning costs and sandbox delays within training runs, and report design trade-offs and operational lessons from production use and comparisons with three open-source systems. We will open-source Sandman's daemon, provider and pool.

cs.DC↗
arXiv2026

Differential Vector Erasure: Unified Training-Free Concept Erasure for Flow Matching Models

Text-to-image diffusion models have demonstrated remarkable capabilities in generating high-quality images, yet their tendency to reproduce undesirable concepts, such as NSFW content, copyrighted styles, or specific objects, poses growing concerns for safe and controllable deployment. While existing concept erasure approaches primarily focus on DDPM-based diffusion models and rely on costly fine-tuning, the recent emergence of flow matching models introduces a fundamentally different generative paradigm for which prior methods are not directly applicable. In this paper, we propose Differential Vector Erasure (DVE), a training-free concept erasure method specifically designed for flow matching models. Our key insight is that semantic concepts are implicitly encoded in the directional structure of the velocity field governing the generative flow. Leveraging this observation, we construct a differential vector field that characterizes the directional discrepancy between a target concept and a carefully chosen anchor concept. During inference, DVE selectively removes concept-specific components by projecting the velocity field onto the differential direction, enabling precise concept suppression without affecting irrelevant semantics. Extensive experiments on FLUX demonstrate that DVE consistently outperforms existing baselines on a wide range of concept erasure tasks, including NSFW suppression, artistic style removal, and object erasure, while preserving image quality and diversity. Code is available at https://github.com/zhiqizhang23/DVE.

cs.CV↗
arXiv2026

ATTUNE: Agentic Throughput Tuning for LLM Training

Training a large language model can occupy thousands of GPUs for weeks, and its speed depends on its recipe: how the model is split across GPUs, how memory is traded for compute, and which kernels and communication overlaps it uses. Finding a good recipe is still manual work for experts. Automated tuners search fixed lists of knobs and check only that a job runs, so they miss large gains and can ship changes that silently alter what the model learns. ATTUNE tunes any working recipe unattended within a GPU-hour budget. A coding agent, guided by lessons from past tuning runs, finds each run's bottleneck and changes one setting at a time, and a correctness gate outside the agent refuses changes known to alter training. ATTUNE is deployed in production at Amazon Ads, a business that reported $19.8 billion in revenue in Q2 2026. Its recipes train the models that power Amazon's Ads Agent. On three of the models it tunes, its recipes raise throughput per GPU by 1.7x on average and up to 3.9x over internal expert-tuned recipes. It also speeds up four of NVIDIA's benchmark recipes by 7-26% and held-out public recipes by up to 2.6x, and beats Bayesian optimization, random search and OptiCo, a recent LLM-agent tuner, at the same GPU-hour budget. We release ATTUNE at github.com/amazon-science/attune and OpenRecipes, a leaderboard of more than 60 verified recipes, at openrecipes.ai.

cs.DC↗
arXiv2026

Too Hard, Too Easy, Just Right: Performance-Driven Challenge Adaptation in Virtual Reality Firefighter Training

Effective firefighter training requires practice under dynamic and stressful emergency conditions, yet real-world live-fire exercises are costly, risky, and difficult to scale. Virtual Reality (VR) offers a safe and controllable alternative, but many existing VR firefighter training systems rely on fixed scenarios or provide limited difficulty adaptation, making it difficult to maintain an appropriate level of challenge for skill development as trainees' performance changes over time. To address this limitation, we present ProACT, a performance-driven adaptive training system that regulates scenario difficulty through closed-loop interactions between trainee performance and simulated fire dynamics. Specifically, ProACT combines a global module that continuously adjusts scene-level fire spread dynamics in real time based on trainee performance with a local module that periodically updates object-level flammability across the environment. Our evaluation shows that ProACT improves both objective training performance and subjective user experience compared with existing adaptive and non-adaptive baselines. We further discuss how adaptive immersive environments can extend beyond firefighting to broader training domains.

cs.AI↗
arXiv2026

Rosemary: Efficient and Adaptable In-Storage Acceleration for Fully Homomorphic Encryption

Fully Homomorphic Encryption (FHE) enables computation on encrypted data and is promising for privacy-preserving computing. However, FHE deployment remains constrained due to massive computational, memory, and data-movement overheads. Existing acceleration strategies often overlook the storage I/O bottleneck caused by the enormous size of FHE data in large-scale applications, as most prior works focus on small benchmarks that fit in main memory. Our studies show that state-of-the-art FHE systems spend 71\% to 85\% of their execution time moving data between storage and compute units rather than performing useful computation. To address the I/O bottleneck, we present Rosemary, a heterogeneous FHE acceleration system that incorporates a novel in-storage processing (ISP) architecture with compiler-level and algorithmic optimizations. Our key innovations include 1) an adaptable hardware-aware FHE engine inside a NAND flash-based solid-state drive (SSD) for efficient and cost-effective in-storage FHE processing, 2) a Throughput-Data-Aware Program Allocator (TDAP) that partitions FHE workloads into heterogeneous programs, and 3) a Compiler-Level Ciphertext-Aware Optimizer (CLCAO) that exploits ISP-friendly FHE kernels for low-cost ISP execution. Across a diverse set of FHE workloads, including encrypted deep learning inference and cloud database queries, Rosemary achieves geometric-mean speedups of $185.7\times$ over state-of-the-art CPU/GPU implementations and $3.7\times$ over prior FHE accelerators, while improving energy efficiency by up to $3.9\times$ and Energy-Delay-Area-Product (EDAP) by up to $16.5\times$.

cs.CR↗
arXiv2026

Second order Recurrences, quadratic number fields and cyclic codes

Wall-Sun-Sun primes (shortly WSS primes) are defined as those primes $p$ such that the period of the Fibonacci recurrence is the same modulo $p$ and modulo $p^2.$ This concept has been generalized recently to certain second order recurrences whose characteristic polynomials admit as a zero the principal unit of $\mathbb{Q}(\sqrt{d}),$ for some integer $d>0.$ Primes of the latter type we call $WSS(d).$ They correspond to the case when $\mathbb{Q}(\sqrt{d})$ is not $p$-rational. For such a prime $p$ we study the weight distributions of the cyclic codes over $\mathbb{F}_p$ and $\mathbb{Z}_{p^2}$ whose check polynomial is the reciprocal of the said characteristic polynomial. Some of these codes are MDS (reducible case) or NMDS (irreducible case).

math.NT↗
arXiv2026

The rank of $\mathbb{Z}_{2^s}$-linear Hadamard codes

The canonical $\mathbb{Z}_{2^s}$-linear Hadamard codes are the Gray images of the $\mathbb{Z}_{2^s}$-additive codes. Their rank, the dimension of the binary linear span, was known in closed form only for $s\le 3$, and for $s \ge 4$ only through computer computations for small lengths. We give a method that computes the rank for every $s$. We show that the rank is a binomial polynomial in the type parameters, with coefficients $ρ_s(\boldsymbol{k})$ independent of the length. We compute all the coefficients for $s\le 5$, which yields explicit rank formulas over $\mathbb{Z}_4$ and $\mathbb{Z}_8$ (recovering the known ones), and over $\mathbb{Z}_{16}$ and $\mathbb{Z}_{32}$.

cs.IT↗
arXiv2026

Chart the Invisible: Interactive Perception for Hidden-Target Exploration in GUI Agents

Despite rapid progress in command-line interface (CLI) tasks, recent LLM-based agents still struggle with reliable task execution through graphical user interfaces (GUIs). A key challenge is the limited content visible within a bounded viewport, especially on mobile devices, where many task-relevant targets remain hidden and can only be revealed through swipe interactions. Therefore, although most GUI agents support swipe as an action primitive, the lack of target visibility makes it difficult for them to decide when, where, and how to swipe to reveal the hidden content. Our quantitative study further shows that GUI agents incorrectly choose a click action in 47.8% of cases that require swipe exploration. To bridge this gap, this paper proposes equipping GUI agents with scrollable-region perception to support hidden-area exploration before action selection. To this end, we trained the first multimodal model for scrollable-region perception on a manually collected dataset of 667 UI screens, called Hagasu. By correctly identifying 78.5% of scrollable regions, experiments across nine GUI agents show that integrating Hagasu improves the mean end-to-end completion rate from 64.8% to 82.7% on explore-dependent tasks. To facilitate follow-up research, we release the swipe dataset and the source code of Hagasu.

cs.SE↗
arXiv2026

Understanding Communication Failures in Multi-Agent LLM Systems

Multi-agent LLM systems aim to improve reasoning through debate, evidence sharing, and mutual correction, but interaction can also introduce failures that final accuracy alone cannot explain. Agents may converge too early, ignore dissent, suppress correct minority answers, or preserve a wrong framing across rounds. We present a corpus-grounded taxonomy of communication failures, built from 3,211 failed multi-round trajectories across medical QA, coding, finance, multi-hop reasoning, ethics, and embodied text tasks. Using this taxonomy, we analyze failure distributions across datasets and system designs, round-level error dynamics, recovery hardness, majority transitions, minority-signal suppression, and self-reported confidence. We find that failures are dominated by recovery/correction and consensus/decision problems rather than isolated first-round mistakes, with many emerging after agents first observe peer responses. Based on these findings, we recommend stronger disagreement handling, evidence verification, minority-signal preservation, and recovery checks before finalization.

cs.AI↗
arXiv2026

WaitLens: Temporal VLM-Based Adaptive Waiting for Industrial GUI Testing Agents

Improving the efficiency of GUI testing agents is important for industrial deployment. Beyond LLM inference, fixed waiting between consecutive interactions introduces substantial overhead. Our industrial study reveals that fixed waiting accounts for 42.3% of the time within the sampled interaction intervals, comparable to the 54.8% spent on model calls. Moreover, 62.9% of the waiting time is spent after the GUI has already finished loading. Although prior studies have proposed adaptive waiting strategies, their reliance on app-specific data or platform-specific information limits their applicability across diverse industrial environments. To this end, we propose WaitLens, which fine-tunes a lightweight VLM on GUI loading trajectories annotated with waiting end times. We introduce temporal resampling to construct diverse training examples, enabling the model to learn varying GUI loading behaviors from temporal observations. Experiments show that WaitLens achieves 86.08% accuracy in deciding whether to keep waiting, compared with 54.13% for the backbone VLM. Our evaluation using recordings from Meituan's industrial testing pipeline shows that WaitLens reduces mean extra waiting after the interface is ready by 94.98% compared with a 20-second fixed wait.

cs.SE↗
arXiv2026

Focused Omni-Modal GUI Agents

Graphical User Interface (GUI) agents increasingly act over streaming omni-modal observations spanning screenshots, audio, video, interaction history, and device states. However, exposing this complete context uniformly at every step ignores that relevance is transient and sparse: the signal determining one action can become clutter at the next. Because executed actions feed back into the context, these errors compound into Sequential Distraction, driving the agent into irreversible trajectory drift. We formulate the underlying requirement as Focused Perception: a representation that is step-adaptive, action-sufficient, and non-intrusive. We realize this in FocusGUI, a lightweight adapter that dynamically prioritizes multimodal channels against a step-specific decision reference. Instead of uniformly processing all inputs, it compresses the most relevant evidence into a compact latent and subtly injects it into the frozen decoder's embeddings, guiding the model's attention without altering its native parameter space. FocusGUI trains from action supervision alone, requiring neither modality-relevance labels nor backbone fine-tuning. Across four omni-modal backbones, it raises overall Type Match by +33.2 and Exact Match by 3.0 times over the frozen backbone on average, and Type Match by +30.1 over the strongest existing enhancement method.

cs.CV↗
arXiv2026

Cache2LoRA: Understanding and Improving Context Internalization

Context-to-LoRA methods encode documents as low-rank parameter updates, allowing language models to answer without retaining the source text in their decoding context. How reliably do these updates preserve facts and support their use under different instructions? Controlled diagnosis of existing context-to-LoRA methods separates fact discrimination, target retention, and response adaptation. Competing sources influence answers, while target-only construction and memory-state optimization improve recovery at fixed update size. Recovering facts also does not ensure that answers follow response instructions. We introduce Cache2LoRA, which uses each query to aggregate evidence from a reusable frozen document cache before generating LoRA updates. Trainable encoding operates only on the query and memory-token suffix, and multi-style supervision teaches both evidence preservation and answer expression. Across 17 tasks, Cache2LoRA improves on existing context-to-LoRA methods and achieves macro F1 of 0.663, approaching full-context in-context learning at 0.671. For 32K documents and 256-token outputs at batch size 16, its decoding throughput is 5.9$\times$ that of full-context ICL. At matched memory size, generated LoRA preserves answer accuracy more reliably than the learned-KV variant under long, diverse intervening text and combines more effectively with retained document KV, yielding higher hybrid F1 and EM.

cs.AI↗
arXiv2026

Joint, Decoupled, or Agent: When Each Multimodal Alignment Paradigm Wins

Jointly pretrained multimodal encoders are a common way to combine modalities, while alternatives integrate their information during downstream training or inference. When does each approach win? We compare three paradigms, instantiated with public checkpoints, across 16 paired datasets in five domains and multiple training-label budgets. JOINT trains a fusion head on frozen, jointly pretrained encoders. DECOUPLED trains a fusion head on frozen, independently pretrained encoders. AGENT gives a frozen LLM task instructions, labelled examples, available sample text, and supervised unimodal readouts. No paradigm dominates the complete 144 dataset-budget combinations: DECOUPLED wins 54, JOINT 44, and AGENT 46 by mean utility. ScienceQA, MELD, and MetMeme shift from AGENT to fusion as supervision increases. Between the fusion paradigms, JOINT leads on four of five QA datasets across all evaluated budgets, while higher unimodal 5-NN accuracy is associated with a relative shift toward DECOUPLED across eight classification datasets (Spearman $ρ=-0.67$). We test whether representation and task diagnostics predict these preferences on held-out datasets before fitting their candidate systems. A two-step cascade combines these diagnostics with task and budget metadata to select AGENT or fusion, then JOINT or DECOUPLED. Under leave-one-group-out evaluation, it selects the winner in 71.4% of cells versus 54.2% for metadata-only routing, with mean relative regret 0.028 versus 0.057. Using geometry, supervised representation diagnostics, and text-label matching together yields higher accuracy and lower regret than using any one of these sources with metadata. These results connect paradigm preferences to supervision budget, task format, and unimodal 5-NN accuracy. Target-task diagnostics thus support selecting among the three paradigms before candidate fitting.

cs.AI↗
arXiv2026

Robust Reinforcement Learning for Dynamic Treatment Regimes under Model and Measurement Uncertainty

We develop a robust reinforcement learning (RL) framework for learning adaptive treatment policies that generalize across patients and remain reliable under errors in the measured biomarker and shifts in the disease dynamics. By modeling the patient as a hidden context, we address generalization and robustness in a unified manner through robust contextual Markov decision processes. Applying our framework to adaptive therapy for metastatic prostate cancer in a simulator calibrated to clinical data, we obtain generalizable and robust policies that substantially outperform the existing adaptive therapy method. In particular, we attain robustness with almost no loss of performance when the biomarker is accurate and the disease evolves as anticipated, and find that the resulting policy Pareto-dominates the existing method across a diverse array of scenarios. Our findings suggest that robust RL holds great promise as a principled methodology for learning reliable adaptive treatment policies.

cs.LG↗
arXiv2026

Clock-Free Stochastic Scheduling for Real-Time Synchronized Content Delivery

Real-time synchronized content delivery is critical for applications such as cloud gaming, collaborative mixed reality, and live streaming, where a server must ensure that a group of users perceives the same content simultaneously for each delivery. Due to heterogeneous and unpredictable network delays, ensuring synchronization is highly challenging. We propose a clock-free scheduling framework that introduces latency compensation to control the server's delivery time for each user, aiming to maximize synchronization performance (i.e., minimizing the gap between the maximum and minimum latencies) while maintaining a low overall latency cost. Unlike prior work requiring synchronized clocks or client-side timestamps, our approach uses only acknowledgment feedback, enabling flexible deployment across diverse environments. We formulate the problem as a convex stochastic optimization with a non-smooth objective. To solve the problem, we employ the stochastic proximal point (SPP) algorithm that updates the delay compensation, using only the previous latency samples, rather than the latency distribution. By exploiting the subproblem structure, we apply ADMM to enable efficient updates with per-iteration complexity of at most $\mathcal{O}(N\log N)$, where $N$ is the number of users. We establish a convergence rate of $\mathcal{O}(1/\sqrt{K})$, where $K$ is the number of content deliveries for transmission. The framework is further extended to handle transmission failures, where packet loss results in successful delivery to only a subset of users. Finally, simulations using both synthetic and real-world latency traces demonstrate the consistent superiority of our algorithm against baseline methods.

cs.NI↗
arXiv2026

Turán problem on the union of three intervals of equal length

This paper addresses the Turán extremal problem on the symmetric three-interval set $Ω_λ= (-1,1) \cup (λ-1,λ+1) \cup (-λ-1,-λ+1)$, $λ\geq 0$. We establish a discrete approximation principle relating the continuous Turán problem on $Ω_λ$ to the discrete Turán problem on an associated finite discrete set. This enables the continuous problem to be exactly expressed as a limit of finite nonnegative trigonometric polynomial problems, which are equivalent to finite semidefinite programs. Using this framework, we compute the Turán constant for integer and rational parameters $λ$. We further analyze the dependence of the Turán constant on $λ$ and derive upper bounds for all $λ> 5$.

math.CA↗