arXiv subjects

Yong Wu

Publications and source records attributed to Yong Wu.

At least 19 recordsLinked to original sources

arXiv2026

NewtonGS: Physics-Structured Object-Level Neural Newtonian Dynamics for Gaussian Scene Animation

Animating objects in a static 3D Gaussian scene requires an explicit object-level dynamic state and a controllable model of object motion. Existing dynamic Gaussian methods primarily reconstruct time-varying scenes or simulate deformation, rather than provide compact object states for direct control. To address this gap, we present NewtonGS, a physics-structured framework for object-level state rollout and Gaussian scene animation. NewtonGS represents each object with a 22-dimensional state covering pose, linear and angular velocity, anisotropic scale and its rate, mass, and contact properties. Its Gaussian Neural Newtonian Dynamics (Gaussian-NND) model combines analytic translation, quaternion kinematics, gravity, damping, and scale-restoration dynamics with learned continuous and contact residuals. A discrete event map handles floor contact. Predicted poses and scales define a shared affine transformation that updates the means and covariances of all Gaussians associated with each object. We construct two procedurally generated datasets: State-32 for state-rollout evaluation and Gaussian-32 for state-to-Gaussian transformation. On both the in-distribution and velocity-range-shift splits of State-32, NewtonGS achieves lower trajectory RMSE, final displacement error, and velocity RMSE than five analytic baselines. Experiments on Gaussian-32 further demonstrate effective conversion from predicted states to animated Gaussian objects.

cs.CV↗

arXiv2026

Plasma screening and configuration interaction effects induced large enhancement on L-shell photoionization cross sections and opacity

An opacity model that incorporates improved treatments of both plasma screening and configuration interaction (CI) effects is proposed, and a 25-30% enhancement on the iron L-shell opacity is predicted at solar interior temperatures. It is originated from the plasma screening induced 14-17% enhancement on the photoionization cross sections and the CI induced 10-20% enhancement on photoexcitation and photoionization cross sections for open L-shell ions. These explain the long-standing discrepancy between theoretical and experimental iron opacity [Nature 517, 56], and the relatively weaker enhancements on chromium and nickel opacity [Phys. Rev. Lett. 122, 235001] due to the sensitivity of these effects to the different L-shell electron population and plasma temperature/density. This letter provides the systematic interpretation of L-shell opacity measurements at solar interior temperatures, and advances the accurate simulation of opacity and radiative transport in high-energy-density plasma.

physics.atom-ph↗

arXiv2026

MOSAIC: Adaptive Inter-layer Composition for Efficient Heterogeneous Vision-Language Models

Vision-Language Models (VLMs) have achieved success using homogeneous Transformers to process multimedia data. Recent studies show that heterogeneous structures interleaving efficient mechanisms, like linear attention, improve both performance and inference latency over homogeneous designs. However, these efforts rely on handcrafted static mixing patterns, which are sub-optimal and difficult to adapt to specific hardware. To bridge this gap, we propose Multi-Objective Search for Adaptive Inter-layer Composition (MOSAIC), a hardware-aware search method that automatically transforms homogeneous models into optimized heterogeneous architectures. MOSAIC integrates diverse efficiency mechanisms--including linear, sparse, and low-rank operators--into a unified search space. By formulating the selection as a multi-objective Mixed Integer Programming (MIP) problem, our method identifies optimal configurations that maximize downstream performance under strict hardware latency constraints. To mitigate performance degradation from structural transitions, we introduce a two-stage parameter recovery process: global off-policy distillation to stabilize internal representations, followed by a dual-teacher on-policy distillation leveraging a 235B oracle for knowledge expansion and the original 4B teacher for distributional stability. We validate MOSAIC through MOSAIC-4B, derived from Qwen3-VL-4B-Instruct. Results demonstrate that MOSAIC-4B matches the baseline's performance across multiple benchmarks while requiring less than 2% of the original training cost. Furthermore, it substantially improves inference efficiency, achieving 1.76x prefilling and 2.54x decoding speedups.

cs.CV↗

arXiv2026

Selective Coupling of Decoupled Informative Regions: Masked Attention Alignment for Data-Free Quantization of Vision Transformers

Data-Free Quantization (DFQ) addresses data security concerns by synthesizing samples, without accessing real data. It has garnered increasing attention in the context of Vision Transformers (ViTs), owing to the superiority of the self-attention mechanism compared to classical convolutional operation. However, previous DFQ arts for ViTs often suffer from a distribution mismatch between synthetic samples and input distribution expected by quantized models Q, resulting in the suboptimal performance. In this paper, we propose a novel Masked Attention Alignment approach for Data-Free Quantization of ViTs, named MaskAQ, revealing that: 1) the semantics in the self-attention mechanism is predominantly localized to a sparse subset of patches, called informative regions; 2) the informative regions dominate the mutual information between synthetic samples and Q's outputs. To these ends, we incorporate differential entropy maximum over patch similarity of synthetic samples, to decouple informative regions from noisy background. To couple with varied Q, the informative regions are selected to align full-precision models with Q via a masked attention alignment objective, thus yielding high-quality synthetic samples. Furthermore, a periodic sample refreshing strategy comes up to endow MaskAQ with the capacity to continually adapt to the evolving state of Q throughout the training process, to preserve desirable mutual information with synthetic samples. Extensive experiments verify the merits of MaskAQ over state-of-the-art approaches across multiple backbones and downstream tasks. Our code is available at https://github.com/hfutqian/MaskAQ.

cs.CV↗

arXiv2026

$C^{\infty}$ regularity of the Alt-Phillips Functional for negative powers

In this paper, we study the regularity of the free boundary for minimizers of the Alt-Phillips functional with negative powers \[\mathcal{E}_{\gamma}(u)=\int_{\Omega}\frac{1}{2}|\nabla u|^2+\frac{1}{\gamma}u^{-\gamma}\chi_{\{u>0\}}dx,\quad\gamma\in(0,2).\] We proved that the free boundaries are $C^{\infty}$ at regular points. A key technical tool is the linearized operator for the PDE satisfied by the partial derivatives of a solution to the Alt-Phillips Euler-Lagrange equation in the negative power case. For this operator we establish a comparison principle, which may have further applications to the Alt-Phillips problem with negative powers.

math.AP↗

arXiv2026

Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks

Rubric-based Reinforcement Learning (RL) has emerged as a promising approach for aligning Large Language Models (LLMs) with complex, open-domain instruction following tasks. However, existing methods predominantly rely on response-level rewards, introducing severe reward sparsity and reward ambiguity problems. To address these issues, we propose Rubrics to Tokens (RTT), a novel rubric-based RL framework that bridges coarse response-level scores and fine-grained token-level credit assignment. RTT introduces a Token-Level Relevance Discriminator to predict which tokens in the response are responsible for a specific constraint, and optimizes the policy model via RTT-GRPO, which integrates response-level and token-level advantages within a unified framework. Furthermore, when transitioning from one-dimensional, outcome-level reward to three-dimensional reward space in the token-level rubric-based RL, we propose a novel group normalization method, called Intra-sample Token Group Normalization, to accommodate this shift. Extensive experiments and benchmarks demonstrate that RTT consistently outperforms other baselines in both instruction- and rubric-level accuracy across different models.

cs.CL↗

arXiv2026

ContextBudget: Budget-Aware Context Management for Long-Horizon Search Agents

LLM-based agents show strong potential for long-horizon reasoning, yet their context size is limited by deployment factors (e.g., memory, latency, and cost), yielding a constrained context budget. As interaction histories grow, this induces a trade-off between retaining past information and staying within the context limit. To address this challenge, we propose Budget-Aware Context Management (BACM), which formulates context management as a sequential decision problem with a context budget constraint. It enables agents to assess the available budget before incorporating new observations and decide when and how much of the interaction history to compress. We further develop BACM-RL, an end-to-end curriculum-based reinforcement learning approach that learns compression strategies under varying context budgets. Experiments on compositional multi-objective QA and long-horizon web browsing benchmarks show that BACM-RL consistently outperforms prior methods across model scales and task complexities, achieving over $1.6\times$ gains over strong baselines in high-complexity settings, while maintaining strong advantages as budgets shrink, where most methods exhibit a downward performance trend.

cs.AI↗

arXiv2026

Radiative Association of Ag and H: Formation of AgH from Ab Initio Calculations

Radiative association processes leading to the formation of AgH in cold astrophysical environments are investigated for the first time using full quantum scattering theory. High accuracy potential energy curves and transition dipole moments for the low-lying electronic states (X$^1\Sigma^+$, A$^1\Sigma^+$, $1^1\Pi$, $3^1\Sigma^+$, $2^1\Pi$) are computed employing the internally contracted multireference configuration interaction method with Davidson correction. Vibrationally and rotationally resolved radiative association cross sections are calculated for transitions from these initial states to the ground X$^1\Sigma^+$ state. Prominent shape resonances arising from quasi-bound rovibrational levels behind centrifugal barriers are identified, with the $2^1\Pi \to$ X$^1\Sigma^+$ channel exhibiting the strongest contribution at low collision energies. Stimulated radiative association under blackbody radiation fields (up to $T = 20\,000$ K) produces modest enhancements, predominantly in the ground-state channel. Thermal rate coefficients computed over 10$^{-1}$--$10^4$~K reveal a general decreasing trend with temperature for all channels. The results provide essential kinetic data for astrochemical models of transition-metal hydride formation in low-temperature interstellar and circumstellar environments.

physics.atom-ph↗

arXiv2026

SkillRouter: Skill Routing for LLM Agents at Scale

Reusable skills let LLM agents package task-specific procedures, tool affordances, and execution guidance into modular building blocks. As skill ecosystems grow to tens of thousands of entries, exposing every skill at inference time becomes infeasible. This creates a skill-routing problem: given a user task, the system must identify relevant skills before downstream planning or execution. Existing agent stacks often rely on progressive disclosure, exposing only skill names and descriptions while hiding the full implementation body. We examine this design choice on a SkillsBench-derived benchmark with approximately 80K candidate skills, targeting the practically important setting of large skill registries with heavy overlap. Across representative dense and reranking baselines on this setting, hiding the skill body causes a 37-44 percentage point drop in routing accuracy. Stronger controls show that the missing signal is body-resident rather than a simple length artifact: body-distilled descriptions recover part of the gap, but remain 7-21 points below direct all-field routing, while a metadata-only encoder trained with the same data remains 14.0 points below its all-field counterpart. Motivated by this finding, we present Skillrouter, a compact 1.2B body-aware retrieve-and-rerank pipeline. Skillrouter achieves 74.0% Hit@1 on our benchmark -- the strongest average top-1 routing performance among the baselines we evaluate -- while using 13$\times$ fewer parameters and running 5.8$\times$ faster than the strongest base pipeline. The ranking gains further generalize to a supplementary benchmark independently constructed from three skill sources. In a complementary end-to-end study across four coding agents, routing gains transfer to improved task success, with larger gains for more capable agents.

cs.LG↗

arXiv2026

Dielectric response and structural properties of finite-temperature electron liquids

The dielectric response and structural properties of finite-temperature electron liquids are central to accurately describing the physical behavior of electronic systems. This study presents a robust analytical model for the static structure factor of the uniform electron gas, combining physically motivated form for the static structure factor with constraints derived from high-accuracy path integral Monte Carlo simulations. The model accurately reproduces key features of the static structure factor across a broad range of temperatures and densities. Using this static structure factor, the density response function is directly evaluated, enabling a self-consistent definition of the static local field correction. As practical applications, the model is employed to investigate the low-velocity stopping power and the electron-ion friction coefficient. Results derived for the friction coefficient show good agreement with simulation data at moderate coupling and degeneracy. The proposed approach provides a computationally efficient and reliable method for characterizing the static response properties of correlated electron systems, facilitating improved simulations of energy deposition and ionic transport in warm dense matter and other strongly coupled quantum plasmas.

physics.plasm-ph↗

arXiv2026

Plasma Screening Effects in Stark Broadening: A Fully Relativistic Close-Coupling Approach

Stark broadening of spectral lines in plasmas is a cornerstone of opacity modeling and plasma diagnostics, with critical implications for controlled fusion and astrophysics. Despite recent advances in fully quantum-mechanical close-coupling calculations for electron-impact broadening, the impact of denser plasma environments remains largely unexplored due to theoretical bottlenecks associated with electron-ion collision processes. Based on our newly developed close-coupling theory for electron-ion collisions in plasmas, which resolves the problem of extracting short-range scattering phase shifts, we introduce a fully relativistic close-coupling approach for the Stark broadening that incorporates plasma screening effects. Systematic investigations of hydrogenic radiators reveal distinct patterns of line broadening dependence on plasma conditions, offering valuable insights for plasma diagnostic applications. Furthermore, we provide a quantum-mechanical interpretation of the screening factor commonly introduced in semi-classical impact theories. This work establishes a robust foundation for future studies on complex atomic systems in high-density plasmas.

physics.plasm-ph↗

arXiv2026

R-Matrix Theory for Electron-Ion Collisions in Plasmas

Electron-atom collisions in warm dense plasmas are crucial for astrophysics and controlled fusion research, where calculating short-range scattering matrices under screening plasma potentials is essential. While electron-neutral atom collisions are tractable using the standard Riccati-Bessel wavefunctions in the asymptotic region, electron-ion collisions face challenges due to the extended range of the screened Coulomb potential, which lacks analytical solutions or numerical code packages for asymptotic regular and irregular wavefunctions. We introduce an R-matrix theoretical framework for general screened potentials and develop a numerical method to compute these asymptotic wavefunctions efficiently. Our approach yields short-range scattering phase shifts that remain invariant with respect to the matching point in the asymptotic region. Applying the Debye screening potential as an illustrative example, we calculate elastic and electron-impact excitation collision strengths for H-like ions (He, C, Ne) across varying temperatures and densities. The calculations show that Debye screening systematically modifies resonance structures and progressively lowers excitation thresholds. Nevertheless, the effective collision strengths and rate coefficients exhibit approximate scaling laws. These findings enable convenient access to electron collision data in plasma environments, advancing plasma diagnostics and modeling.

physics.atom-ph↗

arXiv2026

Boundary-Aware NL2SQL: Integrating Reliability through Hybrid Reward and Data Synthesis

In this paper, we present BAR-SQL (Boundary-Aware Reliable NL2SQL), a unified training framework that embeds reliability and boundary awareness directly into the generation process. We introduce a Seed Mutation data synthesis paradigm that constructs a representative enterprise corpus, explicitly encompassing multi-step analytical queries alongside boundary cases including ambiguity and schema limitations. To ensure interpretability, we employ Knowledge-Grounded Reasoning Synthesis, which produces Chain-of-Thought traces explicitly anchored in schema metadata and business rules. The model is trained through a two-stage process: Supervised Fine-Tuning (SFT) followed by Reinforcement Learning via Group Relative Policy Optimization. We design a Task-Conditioned Hybrid Reward mechanism that simultaneously optimizes SQL execution accuracy-leveraging Abstract Syntax Tree analysis and dense result matching-and semantic precision in abstention responses. To evaluate reliability alongside generation accuracy, we construct and release Ent-SQL-Bench, which jointly assesse SQL precision and boundary-aware abstention across ambiguous and unanswerable queries. Experimental results on this benchmark demonstrate that BAR-SQL achieves 91.48% average accuracy, outperforming leading proprietary models, including Claude 4.5 Sonnet and GPT-5, in both SQL generation quality and boundary-aware abstention capability. The source code and benchmark are available anonymously at: https://github.com/TianSongS/BAR-SQL.

cs.CL↗

arXiv2026

IndexTTS 2.5 Technical Report

In prior work, we introduced IndexTTS 2, a zero-shot neural text-to-speech foundation model comprising two core components: a transformer-based Text-to-Semantic (T2S) module and a non-autoregressive Semantic-to-Mel (S2M) module, which together enable faithful emotion replication and establish the first autoregressive duration-controllable generative paradigm. Building upon this, we present IndexTTS 2.5, which significantly enhances multilingual coverage, inference speed, and overall synthesis quality through four key improvements: 1) Semantic Codec Compression: we reduce the semantic codec frame rate from 50 Hz to 25 Hz, halving sequence length and substantially lowering both training and inference costs; 2) Architectural Upgrade: we replace the U-DiT-based backbone of the S2M module with a more efficient Zipformer-based modeling architecture, achieving notable parameter reduction and faster mel-spectrogram generation; 3) Multilingual Extension: We propose three explicit cross-lingual modeling strategies, boundary-aware alignment, token-level concatenation, and instruction-guided generation, establishing practical design principles for zero-shot multilingual emotional TTS that supports Chinese, English, Japanese, and Spanish, and enables robust emotion transfer even without target-language emotional training data; 4) Reinforcement Learning Optimization: we apply GRPO in post-training of the T2S module, improving pronunciation accuracy and natrualness. Experiments show that IndexTTS 2.5 not only supports broader language coverage but also replicates emotional prosody in unseen languages under the same zero-shot setting. IndexTTS 2.5 achieves a 2.28 times improvement in RTF while maintaining comparable WER and speaker similarity to IndexTTS 2.

cs.SD↗

arXiv2026

FlashInfer-Bench: Building the Virtuous Cycle for AI-driven LLM Systems

Recent advances show that large language models (LLMs) can act as autonomous agents capable of generating GPU kernels, but integrating these AI-generated kernels into real-world inference systems remains challenging. FlashInfer-Bench addresses this gap by establishing a standardized, closed-loop framework that connects kernel generation, benchmarking, and deployment. At its core, FlashInfer Trace provides a unified schema describing kernel definitions, workloads, implementations, and evaluations, enabling consistent communication between agents and systems. Built on real serving traces, FlashInfer-Bench includes a curated dataset, a robust correctness- and performance-aware benchmarking framework, a public leaderboard to track LLM agents' GPU programming capabilities, and a dynamic substitution mechanism (apply()) that seamlessly injects the best-performing kernels into production LLM engines such as SGLang and vLLM. Using FlashInfer-Bench, we further evaluate the performance and limitations of LLM agents, compare the trade-offs among different GPU programming languages, and provide insights for future agent design. FlashInfer-Bench thus establishes a practical, reproducible pathway for continuously improving AI-generated kernels and deploying them into large-scale LLM inference.

cs.AI↗

arXiv2025

Index-ASR Technical Report

Automatic speech recognition (ASR) has witnessed remarkable progress in recent years, largely driven by the emergence of LLM-based ASR paradigm. Despite their strong performance on a variety of open-source benchmarks, existing LLM-based ASR systems still suffer from two critical limitations. First, they are prone to hallucination errors, often generating excessively long and repetitive outputs that are not well grounded in the acoustic input. Second, they provide limited support for flexible and fine-grained contextual customization. To address these challenges, we propose Index-ASR, a large-scale LLM-based ASR system designed to simultaneously enhance robustness and support customizable hotword recognition. The core idea of Index-ASR lies in the integration of LLM and large-scale training data enriched with background noise and contextual information. Experimental results show that our Index-ASR achieves strong performance on both open-source benchmarks and in-house test sets, highlighting its robustness and practicality for real-world ASR applications.

cs.SD↗

arXiv2025

Uncertainty-Preserving QBNNs: Multi-Level Quantization of SVI-Based Bayesian Neural Networks for Image Classification

Bayesian Neural Networks (BNNs) provide principled uncertainty quantification but suffer from substantial computational and memory overhead compared to deterministic networks. While quantization techniques have successfully reduced resource requirements in standard deep learning models, their application to probabilistic models remains largely unexplored. We introduce a systematic multi-level quantization framework for Stochastic Variational Inference based BNNs that distinguishes between three quantization strategies: Variational Parameter Quantization (VPQ), Sampled Parameter Quantization (SPQ), and Joint Quantization (JQ). Our logarithmic quantization for variance parameters, and specialized activation functions to preserve the distributional structure are essential for calibrated uncertainty estimation. Through comprehensive experiments on Dirty-MNIST, we demonstrate that BNNs can be quantized down to 4-bit precision while maintaining both classification accuracy and uncertainty disentanglement. At 4 bits, Joint Quantization achieves up to 8x memory reduction compared to floating-point implementations with minimal degradation in epistemic and aleatoric uncertainty estimation. These results enable deployment of BNNs on resource-constrained edge devices and provide design guidelines for future analog "Bayesian Machines" operating at inherently low precision.

cs.LG↗

arXiv2025

AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention

Vision-Language-Action (VLA) models have shown remarkable progress in embodied tasks recently, but most methods process visual observations independently at each timestep. This history-agnostic design treats robot manipulation as a Markov Decision Process, even though real-world robotic control is inherently partially observable and requires reasoning over past interactions. To address this mismatch, we reformulate VLA policy learning from a Partially Observable Markov Decision Process perspective and propose AVA-VLA, a framework that conditions action generation on a recurrent state that serves as a neural approximation to the agent's belief over task history. Built on this recurrent state, we introduce Active Visual Attention (AVA), which dynamically reweights visual tokens in the current observation to focus on regions most relevant given both the instruction and execution history. Extensive experiments show that AVA-VLA achieves state-of-the-art performance on standard robotic benchmarks, including LIBERO and CALVIN, and transfers effectively to real-world dual-arm manipulation tasks. These results demonstrate the effectiveness of temporally grounded active visual processing for improving VLA performance in robotic sequential decision-making. The project page is available at https://liauto-dsr.github.io/AVA-VLA-Page.

cs.LG↗