Searcharxiv⌕ Search

arXiv subjects

Xinyu Wang

Publications and source records attributed to Xinyu Wang.

At least 73 records · Page 4Linked to original sources
arXiv2026

Soliton gas for the derivative nonlinear Schrödinger equation: continuum dbar-problem, genus reduction and asymptotics

A soliton gas theory for the derivative nonlinear Schrödinger equation is developed, focusing on the Gerdjikov-Ivanov equation and using its gauge equivalence with the Kaup-Newell and Chen-Lee-Liu equations. Starting from the reflectionless inverse scattering problem, we formulate pure N-soliton solutions through a meromorphic Riemann-Hilbert problem and pass to the continuum limit as the discrete spectrum condenses on the planar spectral domains. Under a suitable scaling of the norming constants, the limit yields a compactly supported dbar-problem. For domains admitting a Schwarz function, in particular elliptic domains, this problem reduces to a Riemann-Hilbert problem on the associated mother body. For the elliptic soliton gas, the large-x and long-time asymptotic behaviors are established: the solution decays as x->+infty, approaches an elliptic finite-gap background as x->-infty, and exhibits stratified long-time sectors described by one-, two-, and three-phase Riemann theta functions. A distinctive feature of derivative nonlinear Schrodinger equation is the z->-z symmetry, which induces a quotient reduction of the effective Abelian geometry via lambda=z^2. We further derive a kinetic equation for the effective velocity of the elliptic soliton gas and an Its-Izergin-Korepin-Slavnov type Fredholm determinant representation of the continuum tau-function.

math.AP↗
arXiv2026

Do World Action Models Generalize Better than VLAs? A Robustness Study

Robot action planning in the real world is challenging as it requires not only understanding the current state of the environment but also predicting how it will evolve in response to actions. Vision-language-action (VLA), which repurpose large-scale vision-language models for robot action generation using action experts, have achieved notable success across a variety of robotic tasks. Nevertheless, their performance remains constrained by the scope of their training data, exhibiting limited generalization to unseen scenarios and vulnerability to diverse contextual perturbations. More recently, world models have been revisited as an alternative to VLAs. These models, referred to as world action models (WAMs), are built upon world models that are trained on large corpora of video data to predict future states. With minor adaptations, their latent representation can be decoded into robot actions. It has been suggested that their explicit dynamic prediction capacity, combined with spatiotemporal priors acquired from web-scale video pretraining, enables WAMs to generalize more effectively than VLAs. In this paper, we conduct a comparative study of prominent state-of-the-art VLA policies and recently released WAMs. We evaluate their performance on the LIBERO-Plus and RoboTwin 2.0-Plus benchmarks under various visual and language perturbations. Our results show that WAMs achieve strong robustness, with LingBot-VA reaching 74.2% success rate on RoboTwin 2.0-Plus and Cosmos-Policy achieving 82.2% on LIBERO-Plus. While VLAs such as $π_{0.5}$ can achieve comparable robustness on certain tasks, they typically require extensive training with diverse robotic datasets and varied learning objectives. The evaluation code for the RoboTwin2.0-Plus benchmark is available at: https://robot-robustness.github.io/RoboTwin2.0-Plus/.

cs.RO↗
arXiv2026

Q-Steer: Action-Value Guidance for Molecular Policy Optimization

Oracle-limited molecular optimization gives reward only after a complete molecule is generated, while each rollout requires many local next-token decisions. This delayed-feedback interface makes molecular policy optimization myopic: an optimizer can learn that a molecule was good without knowing which intermediate actions made it good. We introduce Q-Steer, a rollout-time action-value steering primitive for molecular language models. Q-Steer uses an offline-trained and frozen prefix-action value scorer, PAVS-Q, that estimates the downstream reward of taking a candidate next token under a partial SMILES prefix, then adds a normalized value bonus to sampling logits. The optimizer update rule and online oracle budget are unchanged; the claim is fixed-online-oracle performance, not equal total compute. On PMO23 with a fixed 10,000-call online budget, complete factorial studies across two molecular language-model backbones and four optimizers show that Q-Steer improves mean valid-unique score in all eight backbone-optimizer cells, with positive macro mean-score gains between +0.033 and +0.049 and 18-20 task wins per cell. Mechanism controls show that action identity matters: prefix-broadcast values are nearly neutral, while shuffled action values harm performance. These results support Q-Steer as a reusable rollout-time action-value wrapper that improves average molecular optimization reward across optimizer families and policy backbones without changing the online oracle budget.

cs.LG↗
arXiv2026

Hierarchical Search of Tree Tensor Networks for High-Dimensional Data

Tensor network methods provide a scalable solution to represent high-dimensional data. However, their efficacy is often limited by static, expert-defined structures that fail to adapt to evolving data correlations. We address this limitation by formalizing the structural rounding problem for tree tensor networks and introducing a hierarchical search algorithm HIST, which automatically identifies optimized structures with index reshaping for input tree tensor networks. To navigate the combinatorial explosion of the structural search space, HIST integrates stochastic sub-network sampling with hierarchical refinement. This approach utilizes entropy-guided index clustering to reduce dimensionality and targeted reshaping to expose latent data correlations. Numerical experiments on analytical functions and real-world physics applications, including thermal radiation transport, neutron diffusion, and computational fluid dynamics, demonstrate that HIST exhibits empirical polynomial scaling with dimensionality relative to the sampling budget, bypassing the scalability barriers in prior work. HIST achieves compression ratios $2.5\times$ to $100\times$ higher than standard fixed formats such as Tensor Trains and Hierarchical Tuckers (peaking at $1000\times$). Furthermore, HIST discovers structures that generalize effectively: applying a structure optimized for one data instance to a related target data typically maintains compression performance within $10\%$ of the result obtained by performing structure search on that target data. These results highlight HIST as a robust, automated tool for adaptive data representation and high-dimensional simulation compression with tensor network methods.

cs.CE↗
arXiv2026

Weak stability and random mean-field limit of the phase-spatially extended kinetic Cucker-Smale model

We study the weak stability of the kinetic Cucker-Smale (in short, KCS) model in a phase-spatially extended setting, which can be formally derived from the infinite Cucker-Smale model in the mean-field limit. For a bounded Lipschitz communication weight function, we derive finite-time Osgood-type weak stability for measure-valued solutions with exponential velocity tails and finite spatial second moments. Unlike the phase-spatially confined setting, the solution operator to the KCS model is not Lipschitz continuous with respect to initial data. This is due to the unbounded velocity tail and the corresponding absence of a uniform Lipschitz bound for the alignment force. As an application of weak stability, we obtain an i.i.d. sampling consequence: empirical measures generated from independent initial samples converge to the measure-value solution for the corresponding kinetic model in any finite time interval, in expectation.

math.AP↗
arXiv2026

InteractComp: Evaluating Search Agents With Ambiguous Queries

Language agents have demonstrated remarkable potential in web search and information retrieval. However, many search-agent benchmarks assume that user queries are complete and unambiguous. This assumption leaves under-tested a practical failure mode: agents may face ambiguous requests where the intended target cannot be identified without clarification. Yet most agents lack interactive mechanisms during the search process, and existing benchmarks cannot assess this capability. To address this gap, we introduce InteractComp, a benchmark designed to evaluate whether search agents can recognize query ambiguity and actively interact to resolve it during search. Following the principle of easy to verify, interact to disambiguate, we construct 210 expert-curated questions across 9 domains through a target-distractor methodology that creates controlled ambiguity resolvable only through interaction. Evaluation of 17 models reveals striking failure: the best model achieves only 13.73% accuracy despite 71.50% with complete context, exposing systematic overconfidence rather than reasoning deficits. Forced interaction produces dramatic gains, demonstrating latent capability current strategies fail to engage. Longitudinal analysis shows interaction capabilities stagnated over 15 months while search performance improved seven-fold, revealing a critical blind spot. This stagnation, coupled with the immediate feedback inherent to search tasks, makes InteractComp a valuable resource for both evaluating and training interaction capabilities in search agents. The code is available at https://github.com/FoundationAgents/InteractComp.

cs.CL↗
arXiv2026

Harmonia: Algorithm-Hardware Co-Design for Memory- and Compute-Efficient BFP-based LLM Inference

Large Language Models (LLMs) incur substantial memory and computation costs. Prior works reduce FP-INT arithmetic overhead by converting linear-layer activations to block floating point (BFP), but retain FP activations in attention layers due to accuracy concerns. We propose Harmonia, an algorithm-hardware co-design framework that enables BFP representation and computation across both linear and attention layers. Harmonia first explores BFP configurations to balance model accuracy and activation compression. It then combines asymmetric bit allocation with hybrid offline-online outlier smoothing to compress the KV cache from FP16 to 4-bit-mantissa BFP with less than 1% average accuracy loss on LongBench. To efficiently support all-layer BFP computation, Harmonia incorporates a reconfigurable PE unit for mixed data formats and precisions, a real-time FP16-to-BFP converter, and a flexible tiling-aware dataflow that reduces external memory traffic. Evaluations across eight widely used LLMs show that Harmonia achieves 3.84x higher area efficiency, 2.03x better energy efficiency, and 3.08x speedup on average, with maximum improvements of 5.05x, 3.90x, and 4.62x, respectively.

cs.AR↗
arXiv2026

FinSAgent: Corpus-Aligned Multi-Agent RAG Framework for Evidence-Grounded SEC Filing Question Answering

Financial question answering over U.S. Securities and Exchange Commission (SEC) filings requires retrieving and synthesizing heterogeneous evidence dispersed across long, standardized, and highly redundant disclosures. Existing retrieval-augmented and multi-agent systems typically derive retrieval queries directly from the user's question and rank candidates by semantic similarity. Together, these choices create prior-corpus misalignment: a mismatch between model priors and the target filings' structure, terminology, and evidence standards. As a result, query generation misses corpus-specific evidence, while semantic reranking favors topically similar but evidentially invalid false-positive chunks. We propose FinSAgent, an evidence-grounded multi-agent framework that reframes SEC filing QA as corpus-aligned retrieval planning and corrects both ends with a single principle: inject corpus-side conditioning wherever model priors would otherwise dominate. FinSAgent combines (1) role-specialized agents anchored to the mandated 10-K item structure, (2) database-aware query decomposition that conditions each agent's sub-queries on a lightweight, summary-level view of the local corpus, and (3) multi-path retrieval with a learned feature-gated reranker that separates evidential validity from semantic similarity. Across five offline financial QA benchmarks, FinSAgent improves retrieval coverage and answer correctness over strong single-agent and multi-agent baselines; in a three-arm randomized online experiment with 1,000 anonymous user ratings, it also receives higher scores than baselines.

cs.IR↗
arXiv2026

Novel extended inner shadow in images of Johannsen-Psaltis black holes with thin accretion disks

The Johannsen-Psaltis (JP) metric is constructed by introducing deviation parameters into the Kerr metric. The presence of these deviation parameters provides additional degrees of freedom for testing the extent to which astrophysical black holes are consistent with the Kerr paradigm and for examining the validity of the no-hair theorem. Since black hole imaging provides one of the most direct means of probing the properties of astrophysical black holes, a detailed investigation of the imaging characteristics of JP black holes constitutes a natural first step toward employing the JP metric to explore potential deviations from the Kerr geometry in astrophysical black hole systems. Accordingly, numerical backward ray-tracing simulations are employed to conduct a comprehensive investigation of the image structures, intensity distributions, and redshift-blueshift signatures of JP black holes with both closed and non-closed event horizons. The dependence of these imaging characteristics on the deviation parameter is also investigated systematically. Furthermore, the emergence of an extended inner shadow, distinct from the original inner shadow of the black hole, in images of JP black holes with non-closed event horizons motivated a detailed investigation of its physical formation mechanism using both numerical backward ray-tracing simulations and an approximate analytical framework.

gr-qc↗
arXiv2026

From Evidence to Trajectory: Abductive Reasoning Path Synthesis for Retrieval-Augmented Generation Agents Development

Retrieval-augmented generation (RAG) agent development is hindered by the lack of executable ground-truth agent-environment interaction trajectories. Existing datasets provide questions, answers, and evidence, but lack fine-grained supervision for retriever invocation, dynamic planning, and stepwise decision-making. Reinforcement learning offers a potential solution, but often suffers from sparse rewards and cold-start failures when base large language models (LLMs) lack sufficient reasoning capability. Meanwhile, existing data synthesis methods mainly generate post-hoc rationales rather than executable environment-interaction trajectories. In this paper, we propose EviPath, an evidence-anchored reasoning path synthesis paradigm for RAG agent development. EviPath reverse-engineers executable trajectories from question-answer pairs and supporting evidence through three stages: (i) Abductive Subtask Planning, which decomposes questions and plans dependency-aware solution paths; (ii) Faithful Sub-question Answering, which uses supporting evidence as a proxy environment to generate grounded intermediate thoughts and answers; and (iii) Conversational Fine-Tuning, which converts complete trajectories into a dialogue format for supervised fine-tuning. Experiments on widely used question-answering benchmarks show that an 8B model trained on our synthetic corpus significantly and consistently outperforms state-of-the-art baselines, achieving a 14.7% absolute Exact Match gain in open-domain question answering.

cs.CL↗
arXiv2026

AlphaOracle: Oracle bone script decipherment via human-workflow-inspired deep learning

Approximately 3,000 of the 4,500 oracle bone script (OBS) characters remain undeciphered due to fragmentary inscriptions and sparse evidence. Current AI approaches fail to replicate expert workflows that integrate form analysis, contextual semantics, and philological reasoning. We introduce AlphaOracle, a human-workflow-inspired framework that systematizes OBS decipherment using the largest digitized corpus to date. Its multi-stage pipeline comprises: (i) rubbing parsing; (ii) radical-based morphological analysis with diachronic modeling; (iii) contextual retrieval with semantic alignment; and (iv) philological validation against classical sources. Each stage generates explicit, confidence-weighted evidence chains, culminating in interpretable reports for scholarly verification. Across multiple test characters, AlphaOracle's readings strongly agreed with expert interpretations. In a study of 86 domain specialists, it reduced analysis time by 64% and 79% of participants rated it highly useful. Notably, AlphaOracle resolves the character "Lao" as a toponymic or clan designation, offering concrete revisions to Shang administrative and social interpretations. These results suggest that computational methods aligned with philological practice can facilitate OBS research and provide a conceptual reference for studies of other undeciphered scripts.

cs.HC↗
arXiv2026

Uniform-in-time stability and mean-field limit of the Cucker-Smale-type model with noncompact support

We study the uniform-in-time stability and mean-field limit of an infinite Cucker--Smale-type (ICS-type) model in a fully noncompact setting. First, we clarify the main difficulties arising from the original ICS model and its corresponding kinetic Cucker--Smale (KCS) model, when the spatial support is noncompact. In particular, we show that the velocity diameter may remain constant in time, which invalidates the classical approach based on position and velocity diameters. Moreover, we construct a counterexample showing that the original KCS model does not satisfy uniform-in-time stability in the noncompact spatial setting. To overcome these obstacles, we introduce a new ICS-type model whose communication weight depends on a pairwise spatial moment. Under suitable assumptions, we also establish the uniform-in-time stability of the ICS-type model with respect to initial data in the noncompact setting. Finally, we derive the uniform-in-time mean-field limit for the ICS-type model, together with uniform-in-time stability for the corresponding KCS-type model.

math.AP↗
arXiv2026

Constraining the nuclear symmetry energy with electric dipole polarizability and neutron skin characteristics in \texorpdfstring{$^{208}\mathrm{Pb}$}{208Pb} within antisymmetrized molecular dynamics

The electric dipole polarizability $α_D$ and the neutron-skin thickness $ΔR_{np}$ of $^{208}\mathrm{Pb}$ are two powerful and clean probes for constraining the symmetry energy at subsaturation densities. Within the framework of the antisymmetrized molecular dynamics (AMD) model, the width of the strength function and its dynamical origins are understood, and the $α_D$ and $ΔR_{np}$ data favor effective interaction parameter sets with $S_0\approx32$-34 MeV and $L=64$-87 MeV. In addition, our calculations show that the sensitive densities of $α_D$ and $ΔR_{np}$ range from 0.2$ρ_0$ to 0.57$ρ_0$, and the corresponding values of the symmetry energy at the lower and upper ends of this sensitive density region are $S(0.2ρ_0)=10.18\pm 1.10$ MeV and $S(0.57ρ_0)=22.31\pm 1.32$ MeV.

nucl-th↗
arXiv2026

LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editing

Streaming video editing has made rapid progress, yet practical deployment is still limited by two core issues: maintaining stable backgrounds and non-edited regions over time, and achieving the low latency required for real-time interactive scenarios. Meanwhile, recent streaming video generation methods are mostly developed for synthesis and cannot be directly applied to editing due to the strict preservation requirement and region-specific control. In this work, we present a novel streaming video editing framework that performs causal, frame-by-frame editing with strong content preservation and real-time responsiveness. Our key design is a three-stage distillation pipeline that progressively transfers editing capability from a powerful bidirectional foundation model to an efficient unidirectional streaming editor, enabling stable long-horizon edits without sacrificing visual fidelity. To further support real-time deployment, we introduce an AR-oriented mask cache that reuses region-related computation across frames, substantially reducing redundant processing and accelerating inference. Finally, we establish a dedicated benchmark for streaming video editing. Extensive evaluations demonstrate that our method achieves state-of-the-art visual quality among streaming baselines while drastically boosting inference speed to 12.66 FPS, making it suitable for interactive and augmented reality applications.

cs.CV↗
arXiv2026

On exponential convergence of Chebyshev polynomial approximation for multivariate analytic functions

This paper presents a new analysis of the Chebyshev projection for multivariate analytic functions, drawing on pluripotential theory. It is proved that in any downward closed convex polynomial space, the Chebyshev projection achieves the same exponential convergence rate as the best polynomial approximation. This result enables a precise quantification of the exponential convergence rate of the Chebyshev projection. The analysis is then extended to several related topics, including tensorized Chebyshev interpolation, tensor product Gauss--Legendre quadrature, Padua interpolation and cubature, and Chebyshev-Galerkin method, with the corresponding exponential convergence rate established in each case. Supporting numerical experiments are provided to validate the theoretical results.

math.NA↗
arXiv2026

Tensor Network Structure Search Via Canonical Dimension Tree Enumeration

Tensor networks provide a powerful framework for compressing multi-dimensional data. The optimal tensor network structure for a given data tensor depends on both data characteristics and specific optimality criteria, making tensor network structure search a challenging problem. Existing solutions typically rely on sampling and compressing numerous candidate structures; these procedures are computationally expensive and therefore limiting for practical applications. We address this challenge by decoupling topology enumeration from rank assignment search. We first represent the search space using canonical dimension trees, which encode potential network topology through nested index partitions and inherently rule out redundant and suboptimal topologies by construction. To mitigate the assessment bottleneck, we introduce a mechanism powered by the precomputation of a singular value map. By archiving the singular values of all feasible tensor matricizations, we transform the evaluation of any candidate dimension tree into a constraint-solving problem. This formulation yields an empirically near-optimal rank assignment via simple metadata lookups, allowing us to compute structural costs and bypass expensive tensor decompositions for all but the final selected candidate. Experimental results show that our approach accelerates the structure search by up to 10x while achieving highly competitive compression ratios, outperforming standard tensor trains and hierarchical tuckers by up to 10x, and matching or exceeding state-of-the-art structure search tools. Notably, our approach scales to larger tensors that are unattainable by prior work. Furthermore, the discovered topologies generalize well to similar data; they achieve compression ratios up to 2.4x better than tensor trains or hierarchical tuckers, while maintaining a search time of approximately 110 seconds for 6D tensors of 1-2GB disk size.

cs.CE↗
arXiv2026

MATCH: Modulating Attention via In-Context Retrieval for Long-Context Transformers

The quadratic computational cost of traditional attention mechanisms poses a major bottleneck to the scalability and practical deployment of large language models (LLMs), particularly in long-context scenarios. To improve efficiency, existing approaches often enforce rigid structural constraints such as local attention windows. However, these strategies typically lead to substantial performance degradation on tasks requiring precise long-range recall. In this work, we propose MATCH, a scalable and efficient framework that augments sparsified attention mechanisms with dynamically integrated in-context information through an efficient retrieval system. Empirical results show that MATCH significantly improves the performance of sparse-attention models on both synthetic and real-world natural-language tasks. These findings highlight the versatility of MATCH as a general approach for enhancing in-context retrieval capabilities while maintaining the efficiency benefits of sparse attention architectures.

cs.CL↗
arXiv2026

GeoEdit: Geometry-Aware Object Editing via Dual-Branch Denoising

Precisely manipulating objects in a single photograph (translation, rotation, scaling) while obeying 3D physical constraints remains unsolved for diffusion-based editors. Current 2D methods lack spatial awareness and produce perspective violations. Forcing structural proxies into the latent space also disrupts variance homogeneity, and the resulting self-attention leakage leads to ghosting and background blur. The core difficulty is asymmetric: the relocated object must follow a rigid geometry, yet the uncovered background needs freedom to synthesize plausible content. We present GeoEdit, a training-free Lift-Manipulate-Render-Denoise pipeline that satisfies both constraints. We decouple scene and object in 3D, align them through point correspondence, and render a geometry-aligned proxy with a structural depth map. A Dual-Branch Denoising stage then refines this proxy: a video diffusion backbone preserves object identity, while 3D constraints are injected into the foreground within a narrow denoising window at matching noise variance (variance-homogeneous injection). The background denoises freely. Because the injected signal matches the native latent statistics, self-attention stays undisturbed. We also introduce GeoEditBench, a pose-aware benchmark covering object translation, object rotation, and camera movement with pose-aware evaluation metrics. Experiments confirm consistent gains in geometric accuracy, identity fidelity, and background quality. Our codes are available at https://github.com/Heey731/GeoEdit.

cs.CV↗