arXiv subjects

Xu Zhang

Publications and source records attributed to Xu Zhang.

At least 37 records · Page 2Linked to original sources

arXiv2026

OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use

Computer use has reached a turning point: the focus is expanding from GUI-only interaction to systems that complete tasks on users' computers through visual and programmatic interfaces. We define a generalist computer-use agent (GCUA) as a system in which the model chooses how to observe, act, and organize its work across visual and programmatic interfaces. Yet frontier implementations, such as GPT-6 Astra's computer-use system, remain largely closed, limiting systematic study of their design. We introduce Open Generalist CUA (OpenGCUA), an ongoing project to develop and open the full stack for GCUAs with the community. This report presents the first release of OpenGCUA. Code execution is a key feature of our agent design: it lets intermediate results determine subsequent operations and enables models to apply their existing programming capabilities directly to computer interaction. We also propose harness-to-model decision transfer, moving selected observations and task-state decisions from fixed harness rules to the model. On OSWorld-v2, our full configuration improves task scores across all eight evaluated models, including GPT-6 Astra and Claude Opus 5, raising mean partial score by +19.6 points and reducing reported steps per task by an average of 53.0% relative to batched GUI-only execution. We release our implementation and prompts, evaluation trajectories from all eight models, and preliminary versions of a real-device harness and a cross-device SDK. The harness addresses real-device challenges not fully captured by benchmarks, while the SDK provides a common programming interface across computers, mobiles and browsers. Across four real macOS tasks under our real-device harness, selected Qwen3.8-Max runs use 19.4 times as many reported reasoning tokens as GPT-6 Astra, highlighting substantial differences between models in real-device execution efficiency.

cs.CV↗

arXiv2026

From Documents to Spans: Span-Centric Learning for LLM-based Explainable Medical Coding

Medical coding, which assigns standardized ICD codes to clinical documents, is a fundamental yet challenging task in modern healthcare. It requires extensive knowledge of a large code space and the ability to identify and interpret sparse supporting evidence in long, noisy clinical documents. Most existing methods rely on document-level supervision, which provides limited guidance on supporting evidence and insufficient training signals for rare codes. To address these limitations, we propose Span-Centric Learning (SCL), which separates coding knowledge acquisition from its application in full documents. SCL first learns coding knowledge from short spans, then learns to use this knowledge in full documents from a small set of annotated data. Experiments show that, our approach enables an 8B LLM to surpass state-of-the-art methods using only 248 complete documents for training. It also provides explicit supporting evidence for each predicted code, with strong alignment to human annotations. Further analysis shows that the model can learn knowledge of codes absent from documents through spans and apply that knowledge to full documents.

cs.CL↗

arXiv2026

Look Before You Restore: Learning to Predict Before Acting for Agentic Image Restoration

Agents for mixed-degradation image restoration typically learn what a tool does to an image only after running it: failed calls waste computation and harmful outputs go unnoticed until too late. We propose PreActIR, an agentic restoration framework that predicts tool consequences before execution. For each candidate tool, a tool-conditioned world model forecasts changes in residual degradation, image quality, and fidelity, together with the risk of harm and the uncertainty of the prediction. The agent executes the top-ranked candidate, and an independent verifier commits or rolls back its output before the agent replans from the observed state. PreActIR also audits whether performance is limited by the selector or by the candidate space, and studies targeted tool adaptation as a controlled way to expand that space. In a controlled evaluation that fixes the candidate pool, verifier, and budget and employs reference-annotated degradation states, PreActIR uses 42.5% fewer real tool calls and 63.4% fewer rejected calls than reactive, execute-then-evaluate trial and error, while improving Y-PSNR by 0.92 dB.

cs.CV↗

arXiv2026

MAI-Gym: An Open All-in-One Platform for GUI Agent Evaluation and RL Training across Mobile and Computer Environments

Online reinforcement learning (RL) with verifiable rewards has driven rapid progress in mathematics, coding, and tool use, but scaling this success to GUI agents remains challenging due to the high cost of interactive environments and the scarcity of realistic training tasks. We introduce MAI-Gym, an open all-in-one platform that unifies GUI agent evaluation, data collection, and online RL training across mobile, computer, and web environments. Through a shared rollout pipeline, MAI-Gym ensures consistent agent behavior across evaluation and training, supports ten established benchmarks, and scales to 20,000 parallel environment instances. Beyond scalable infrastructure, MAI-Gym enables automated synthesis of diverse GUI applications and challenging tasks, including resettable mocked apps and coherent cross-app worlds with execution-validated task verifiers. Using MAI-Gym, we train MAI-Gym-9B, a unified 9B agent for mobile and computer use, achieving state-of-the-art performance among open GUI agents under 10B parameters: 78.4% on MobileWorld and 76.2% on OSWorld-Verified. Importantly, RL in a mocked train-ticketing app transfers to its real-world counterpart, improving task success on a phys- ical phone from 28.8% to 48.1%. We release the framework, synthesis skills, mocked apps, 2,000 verified tasks, and the model checkpoint to support repro- ducible research and further advances in scalable GUI agent training.

cs.AI↗

arXiv2026

AIPOCH Open-Science: An Open-Source AI Research Workbench with Reproducible .science Packages

AI-assisted research leaves instructions, computations, artifact revisions and checks whose relationships matter when another researcher receives the work. We present AIPOCH Open-Science, an open-source research workbench with portable .science session packages, and specify an exchange contract for these recorded relationships. Version and assessment targets remain attributable while receiving-workspace identifiers can be remapped, eligible contents can be omitted explicitly, and imported history remains separate from a writable fork. A controlled session tested an incorrect first version, a reviewed correction, an unreviewed revision and a planned but unexecuted step. We compared complete and partial exports with source records extracted independently of the exporter, then inspected imports into separate projects within the same installation. Both conditions retained the tested version and review associations, kept the unreviewed revision distinct from earlier assessments and represented excluded content as unavailable. A documentation-rich ordinary directory also supplied evidence for all seven static inspection questions. A separate registry case connected a table to its recorded task, execution and review; offline execution of its retained script and fixed input reproduced two saved CSV tables byte for byte. The study contributes an implemented exchange contract and a controlled examination of its preservation rules, content-selection constraints and interpretation boundaries.

cs.SE↗

arXiv2026

Extension of Sobolev functions on balls in infinite dimensions

We prove the existence of a bounded Sobolev extension operator $E:W^{p,1}\left( B,P \right) \rightarrow W^{p,1}\left( \ell^{2} ,P \right)$ using a completely new method, where $B\subset \ell^{2}$ is the unit ball and $P$ is any non-trivial centered Gaussian measure on $\ell^{2}$. This solves an open problem posed in the literatures.

math.FA↗

arXiv2026

Evolving Creatives with Clicks: CTR-Driven Advertising image Editing via Attribute Disentanglement and Preference Alignment

The optimization objective for e-commerce advertising creatives has shifted from aesthetic quality to online business metrics like click-through rate (CTR). Recent methods generate appealing ad images from product images by constructing CTR-based reward signals and refining prompt generation via Reinforcement Learning. Such paradigms equally penalize all attributes within an underperforming image, even though some are actually beneficial to clicks. In this work, we propose Evolving Creatives with Clicks (ECC), which reformulates creative optimization as fine-grained editing rather than generation from scratch. ECC selectively refines attributes that hurt CTR while preserving those that contribute positively. Specifically, faced with a rich attribute space across massive images, ECC first identifies 28 attributes significantly correlated with CTR via ANOVA and Random Forest regression. Based on attribute contrasts between images with disparate CTRs, we instruction-tune a prompt model to generate editing suggestions, then refine it with real click preferences since reference-derived suggestions are not uniformly beneficial. Furthermore, we introduce Consistent Preference Optimization (CPO) as ECC's preference optimization strategy, which enforces consistency across semantically equivalent shuffled suggestions to mitigate DPO overfitting and improve preference generalization. Extensive experiments demonstrate that ECC significantly improves the CTR of ad images, outperforming state-of-the-art models by a substantial margin. The complete data pipeline, including attribute extraction and analysis, the resulting dataset, and source code will be released upon acceptance.

cs.CV↗

arXiv2026

A Palais-Smale type condition for optimal control problems

We consider an optimal control problem with a general control constraint and a fixed initial state, for which convexity assumptions are absent. To establish the existence of optimal pairs, we introduce a control version of the Palais-Smale type condition adapted from critical point theory and prove that optimal pairs exist under this condition (see Theorem 2.1). Subsequently, we verify that a specific class of optimal control problems satisfies our Palais-Smale type condition (see Proposition 2.1). Furthermore, as an application, we present an example (Example 3.1) and show that, to the best of our knowledge, no existing result can be applied to establish the existence of an optimal pair in this case (see Section 3.1.1). Example 3.2 shows that finitely many Pontryagin maximum principle candidates do not guarantee the existence of an optimal control. Finally, Example 3.3 demonstrates that the proposed Palais-Smale type condition is sufficient, but not necessary, for the existence of optimal controls.

math.OC↗

arXiv2026

Error-Propagation Modeling for Failure Attribution in LLM-Based Multi-Agent Systems

LLM-based multi-agent systems (MASs) are increasingly used to solve complex tasks through coordinated reasoning, tool use, and interaction with external resources. However, attributing failures in such systems remains challenging because the observed outcome often does not directly reveal the error responsible for the failed execution. In this work, the attribution target is the decisive error, defined as the agent--step pair whose correction would recover the failed execution. Existing approaches largely identify suspicious steps without explicitly modeling how errors propagate across interactions or persist in unresolved loops, making decisive errors difficult to distinguish from downstream failure symptoms. We propose \textbf{E}rror-Propagation \textbf{M}odeling for \textbf{F}ailure \textbf{A}ttribution (\textbf{EMFA}). EMFA constructs a structured representation of the failed trajectory, models both cascading propagation and persistent interaction loops, and uses propagation-aware candidate screening followed by counterfactual verification to identify the decisive agent--step pair. On the Who\&When benchmark, EMFA achieves state-of-the-art step-level attribution accuracy and remains competitive at the agent level. It improves the previous best step-level results by 3.45 and 4.40 percentage points on the Hand-Crafted and Algorithm-Generated subsets, respectively.

cs.AI↗

arXiv2026

Recompose and Refine Latent Reasoning Flows for Vision-Language-Action Models

Latent reasoning enables vision-language-action (VLA) models to transform multimodal observations into task-relevant internal states before generating continuous robot actions. While existing methods learn to generate or refine such states for each policy query, they discard successful reasoning after execution and therefore reconstruct similar computation from scratch. We present Reasoning and Flow Memory (FLOWMEM), a unified VLA model that turns successful latent computation into reusable reasoning experience. Rather than appending a fixed retrieved context, FLOWMEM dynamically retrieves and recomposes compatible latent fragments as the embodied context evolves, forming a reasoning route that follows the temporal structure and progress of successful computation. The route is then refined using current visual and proprioceptive evidence before it conditions action generation. Experiments on RoboMME and LIBERO-Plus show that FLOWMEM attains 48.0% and 77.3% success, outperforming memory-free policies by 1.7 and 4.1 percentage points, respectively. These results demonstrate the value of reusing successful latent computation for closed-loop VLA control.

cs.AI↗

arXiv2026

Geometry-Conforming Finite Element Methods for Interface Problems on Fitted and Unfitted Meshes

We develop an arbitrary-degree geometry-conforming finite element (GC-FE) framework for two-dimensional elliptic boundary value and interface problems on curved domains. Using the Frenet--Serret transformation, curved-boundary and interface-fitted segments are represented exactly, while polynomials in Frenet coordinates generate generally nonpolynomial local shape functions in physical coordinates. For interface-unfitted meshes, GC-FE spaces on curved-boundary elements are coupled with geometry-conforming immersed finite element (GC-IFE) spaces on interface-cut elements, with standard polynomial spaces used elsewhere. We establish optimal approximation, inverse, and trace estimates for the GC-FE spaces. For fitted meshes, we prove well-posedness and optimal error estimates in energy and $L^2$ norms for a symmetric interior penalty discontinuous Galerkin discretization. By retaining the prescribed curves exactly, the method avoids the geometric variational crime associated with curved-geometry approximation and requires no corresponding geometric consistency estimates. Numerical experiments confirm the predicted rates, show global accuracy comparable to nodal isoparametric finite elements and smaller true-interface trace errors in the reported tests, and demonstrate the coupled GC-FE-GC-IFE method on interface-unfitted meshes.

math.NA↗

arXiv2026

Origin of superconductivity in bilayer nickelates: a Quantum Monte Carlo study for a sign-problem-free effective model

Using determinant Quantum Monte Carlo, we investigate the interplay between doping, inter-layer tunneling and onsite Hund's coupling in stabilizing superconductivity (SC) in a two-orbital model for the bilayer nickelate $\text{La}_3\text{Ni}_2\text{O}_7$. With realistic dispersion and for certain values of the interaction parameters, the auxiliary-field-decoupled fermion Hamiltonian has Kramers anti-unitary symmetries which guarantee the absence of a sign problem. The same anti-unitary symmetries can also be used to show there is a second instability towards $(π,π)$ inter-layer bond order (BO) in the strong interaction limit. We indicate the possible connection between this bond order and the enigmatic density wave state observed in experiment, and clarify the decisive role played by the inter-layer tunneling in the competition between SC and BO. Finally, possible directions on how to enhance the SC transition temperature and stabilize the SC phase are also discussed.

cond-mat.str-el↗

arXiv2026

Sobolev spaces in infinite dimensions

The classical theory of Sobolev spaces in finite dimensions is well established. Because infinite-dimensional spaces possess inherent analytical and topological complexities, developing a theory of Sobolev spaces for functions of infinitely many variables---rather than merely extending the classical finite-dimensional theory---is far from routine and has led to long-standing open problems. In this paper, we establish such a theory and systematically determine the extent to which these classical results can be carried over to this setting. By carefully adapting the tools introduced in our recent works, we establish a series of infinite-dimensional counterparts of the classical theorems and, in the process, reveal new phenomena with no finite-dimensional analogue. The methods and concepts developed here, together with the results obtained, furnish a robust framework for further study of Sobolev spaces and related problems in infinite-dimensional analysis.

math.FA↗

arXiv2026

Topology-Consistent Task Planning over Cellular Workflow Complexes for LLM-based Agents

Task planning for LLM agents requires workflows that satisfy both user intent and complex sub-task dependencies. While existing planners work well for sequential or directed acyclic graph (DAG)-like structures, they struggle with workflow patterns such as verification-correction loops, convergent branch merging, and reusable intermediate states that arise naturally in real-world tool orchestration. We present TopoPlanner, a topology-consistent planning framework that lifts tool dependency graphs into cellular workflow complexes and uses them as topologyaware context for LLM tool planning. TopoPlanner retrieves a request-relevant closed subcomplex through cosheaf-consistent cellular retrieval, performs multidimensional structural reasoning over the retrieved topology, and interfaces the resulting cellular representation with the planner LLM for tool-sequence generation. Experiments on four tool-planning benchmarks with topology-guided loop, merge, and loop-merge workflows show consistent improvements over prompt-based and graph-enhanced baselines across different local LLM backbones.

cs.AI↗

arXiv2026

CoMemBench: Benchmarking Collaborative Memory Boundaries across Multi-Agent Workflow Topologies

Multi-agent workflows require task-relevant information to be shared across agents, while irrelevant, stale, unverified, or incompatible information must remain isolated. We call this task-conditioned scope of information a collaborative memory boundary. Workflow topology determines which intermediate artifacts are applicable to which downstream workers and when they cease to be valid, thereby providing a structural stress dimension for sharing and isolation. Existing memory benchmarks primarily evaluate retention and retrieval, whereas multi-agent benchmarks emphasize coordination and end-to-end completion, leaving topology-conditioned memory boundaries largely unmeasured. We introduce CoMemBench, an execution-grounded benchmark for collaborative memory sharing and isolation across multi-agent workflow topologies. It constructs 800 composite workflows across four domains from source-grounded dependency graphs, with node-local specifications, verifiable artifact handoffs, native evaluators, and matched isolation challenges. CoMemBench measures workflow completion, verified node progress, required-handoff reliability, isolation robustness, and token cost. Experiments reveal a sharing-isolation trade-off: broader context improves information availability but can weaken isolation, while system rankings shift across topologies and artifact violations.

cs.AI↗

arXiv2026

GARDiff: Graph-Aligned Residual Diffusion for Probabilistic Multivariate Time-Series Forecasting

Diffusion models have recently shown strong potential for probabilistic multivariate time-series forecasting by modeling complex conditional distributions. Recent decoupled diffusion frameworks further separate forecasting into deterministic prediction and stochastic residual generation, making it natural to derive dependency graphs from deterministic representations and use them to guide residual diffusion. However, we show that this direct structural transfer is unreliable. Although deterministic-derived graphs encode useful global dependency priors, they exhibit substantial edge-level misalignment with residual dependency structures, introducing inaccurate or redundant conditions during residual generation. This reveals a previously overlooked deterministic-to-residual structural alignment problem in decoupled diffusion forecasting. To address this problem, we propose GARDiff, a Graph-Aligned Residual Diffusion framework for probabilistic multivariate time-series forecasting. Instead of treating deterministic-derived graphs as fixed diffusion conditions, GARDiff progressively adapts them to residual generation. Specifically, GARDiff estimates residual uncertainty to distinguish high- and low-uncertainty regions, enabling uncertainty-aware structural refinement, and further performs timestep-aware edge sparsification during reverse diffusion to evolve graph conditions from broad dependency aggregation to localized residual refinement. Extensive experiments on six real-world benchmarks demonstrate that GARDiff consistently improves probabilistic forecasting performance and uncertainty calibration over strong baselines.

cs.LG↗

arXiv2026

Topological Coherence for Self-evolving Multi-agent Systems

Complex tasks inherently couple workflow structure, agent responsibility, collaboration, and memory access: task regions delimit responsibility and tool scope, cross-region dependencies give rise to handoffs, and ownership boundaries delimit private and selectively shared memory. Existing methods can jointly optimize agent and communication structures, yet such optimization does not by itself require responsibility, handoff, and memory boundaries to remain consistent with task dependencies. We term this requirement topological coherence. We introduce TOCOMAS, a Topology-Coherent Multi-Agent System. TOCOMAS grounds a task graph in tool interfaces, organizes compatible task nodes into reusable responsibility domains, and derives dependency-induced and profile-conditioned collaboration together with boundary-regulated memory visibility. During online self-evolution, TOCOMAS proposes coupled changes to agent, collaboration, and memory policies, retaining for subsequent tasks only candidates that satisfy structural constraints and improve evaluated reward. Across BBEH, WorkBench, SWE-Bench-Verified, and CoMemBench, TOCOMAS improves task success over baselines across backbones. CoMemBench also shows gains over the self-evolving baseline in verified progress, handoffs, and memory isolation.

cs.AI↗

arXiv2026

Verifiable constraint qualifications for infinite-dimensional optimization problems and their applications to optimal control problems

This paper employs a finite codimensionality condition to establish an enhanced Fritz John condition for general constrained nonlinear infinite-dimensional optimization problems. In applications, this approach provides a new, unified framework for deriving first-order necessary conditions across a broad class of optimal control problems involving both deterministic and stochastic systems. Compared to existing constraint qualifications, our finite codimensionality condition, which is equivalent to the validity of certain \textit{a priori} estimates, yields a direct and analytically tractable verification method for each application. Furthermore, the core ideas of this method can be further extended to investigate the KKT conditions for infinite-dimensional optimization problems.

math.OC↗