arXiv subjects

Qiang Wang

Publications and source records attributed to Qiang Wang.

At least 19 recordsLinked to original sources

arXiv2026

A Multidimensional Evaluation Framework for Artificial Intelligence-Based Ocean State Forecasts

Artificial intelligence (AI) based ocean state forecasts now approach or surpass those produced by operational ocean forecasting systems as measured by conventional pointwise metrics. However, these metrics may favor spatially smoothed forecasts and cannot determine whether forecast variables remain physically coherent. Forecast evaluation should therefore move beyond overall pointwise accuracy toward assessing forecast trustworthiness. We propose a multidimensional framework comprising overall pointwise accuracy, physical consistency, and distributional consistency. Overall pointwise accuracy summarizes forecast and reference agreement over all valid samples; physical consistency evaluates dynamical balances using model outputs alone; and distributional consistency evaluates distribution tails and multiscale spatial structures. Three AI models were deployed as representative testbeds in a unified operational environment. This comprehensive assessment shows that lower overall pointwise errors cannot preclude substantial dynamical imbalances and suppressed fine-scale variance. Conversely, higher pointwise errors may arise from spatial displacement of coherent structures rather than a poor representation of their amplitude. The results demonstrate that the evaluation of AI ocean forecasting models should be based on the multi-dimensional assessment of forecast trustworthiness, rather than solely relying on a single metric. The study provides a preliminary framework for AI model intercomparison, which is also informative for targeted model refinement and operational implementation.

physics.ao-ph↗

arXiv2026

ASCoR: Asymmetric Semantic-Spatial Learning for Remote Sensing Domain Generalization

Source-only domain generalization (DG) for remote sensing semantic segmentation seeks reliable predictions in unseen domains without target-domain data during training. Although vision foundation models offer transferable knowledge, category recognition remains sensitive to source-specific visual cues, while task-specific adaptation may weaken pretrained transferability. We propose ASCoR, an asymmetric semantic-spatial framework built on SAM3. Its text-conditioned semantic branch uses class-name prompts to provide explicit category guidance. Probe-Once Dynamic Adaptive LoRA distributes a fixed adaptation budget across encoder blocks to limit source-specific specialization. A complementary task-specific dense branch learns multi-scale spatial representations using a separate frozen SAM3 image encoder. A lightweight reliability router integrates the two branches through bounded spatial correction based on local prediction reliability. All components are trained using source-domain supervision only. Experiments on the ISPRS, LoveDA, and OpenEarthMap benchmarks cover nine cross-domain transfers. ASCoR achieves a state-of-the-art mean IoU of 68.61% across six ISPRS transfers, outperforming the previous best method by 2.69 percentage points. It also achieves the best results on all three LoveDA/OpenEarthMap transfers, with 59.07%, 47.95%, and 49.38% mIoU on Rural-to-Urban, Urban-to-Rural, and Urban-to-OpenEarthMap, respectively. Ablation studies, feature visualizations, and spatial reliability analyses support the effectiveness and complementarity of the proposed components. Code: https://github.com/WangZitong42/ASCoR.

cs.CV↗

arXiv2026

What Transfers from Text to Vision? Capability Scaling Laws and Transfer Dynamics for VLMs

Choosing the right large language model (LLM) backbone is the most consequential decision when building a vision-language model (VLM), yet it remains fundamentally unprincipled: compute-based scaling laws fail to generalize across model families, and no framework exists for directly predicting VLM performance before training begins. We propose the Capability-Driven Multimodal Scaling Law, the first cross-family framework that predicts VLM benchmark accuracy from directly observable textual capability. Given a low-dimensional capability score $S$ extracted from LLM textual benchmarks via PCA, we model VLM performance as a function of $S$, with a per-backbone transfer rate and an absorption rate that quantifies data-scaling efficiency. To fit and validate the framework, we train over 150 VLMs on 34 LLMs spanning 7 model families under a strictly controlled recipe. Evaluations on more than 200 textual and 50 multimodal benchmarks show that the law accurately extrapolates transfer rate from models up to 8B parameters to 72B-scale backbones, predicts full VLM training trajectories with high fidelity, and generalizes to entirely held-out model families. Beyond the scaling law, our analysis surfaces actionable insights: certain textual benchmarks negatively correlate with multimodal performance, exposing latent benchmark-gaming behavior; base LLMs outperform instruction-tuned counterparts as VLM backbones due to higher absorption rates and lower data-scaling decay; and different model families occupy distinct positions in the transfer--absorption space. The framework turns backbone selection from costly empirical sweeps into a principled, quantitative decision. Project website: https://wangq-dev.github.io/CDMScaling-Website ; Code: https://github.com/wangq-dev/CDMScaling ; Model weights: https://modelscope.ai/profile/wangq2k

cs.CL↗

arXiv2026

Double-Shock Radiation in Supernova Wind Breakout and EP260321a

We study the supernova (SN) shock breakout (SBO) from an optically thick stellar wind. The SN ejecta-wind interaction forms a forward shock (FS) plus reverse shock (RS) structure. The RS's internal energy and shock velocity are both a fraction $μ\lesssim1/6$ of the FS's. The lower RS velocity makes it easier to build up local thermal equilibrium (LTE) and generate blackbody (BB) emission. In the same time a $μ^{-1}\gtrsim6$ times larger non-BB emission may be produced by the likely non-LTE FS. We apply the theory to analyze the recent detected type Ic-BL SN2026gzf associated X-ray flash EP260321a, and find it is naturally accounted for by a SN wind breakout: the $\sim0.13$\,keV, thermal SBO emission can be produced by the RS during the interaction of a typical type Ic-BL SN ejecta ($E_k\sim10^{52}$erg and $M_e\sim1M_\odot$) and a stellar wind with a mass $M_w\sim10^{-4}M_\odot$ confined within a radius $r_w\sim10^{13}$cm; the X-ray rise time is mainly determined by the RS emission time, whereas the long duration and decrease time are due to the scatterings of SBO photons by a distant shell-like medium which has a large photosphere, $r_{ph}\sim10^{14}$\,cm; the early, within 30\,ks, optical emission can be naturally explained by the cooling emission of a shocked pure C/O envelope, since the observed SBO onset is delayed by $\gtrsim$ a few ks due to the large medium photosphere. BB+non-BB components from RS and FS, respectively, may be common in SN wind breakouts. SBO observations can be crucial in the study of progenitor stars by providing information for the ejecta from the SN explosion and the nearby wind from the pre-SN activities.

astro-ph.HE↗

arXiv2026

PixelSR: Pixel-Space Orthogonal Flow Decoupling for Real-World Image Super-Resolution

Latent diffusion models have recently achieved remarkable progress in Real-world Image Super-Resolution (Real-SR). However, existing methods consistently exhibit a persistent seesaw effect: enhancing perceptual realism compromises structural fidelity. This effect, we argue, is not intrinsic to diffusion models, but largely stems from the architectural constraints of prevailing two-stage frameworks. To address this challenge, we propose PixelSR, a novel single-stage Pixel Diffusion Transformer for Real-SR that operates directly in pixel space. Specifically, PixelSR consists of two core designs: Pixel-space Conditional Diffusion and Orthogonal Flow Decoupling. The former establishes an effective pixel-space restoration framework by integrating spatially aligned LR conditioning, visual semantic guidance, and lightweight texture refinement, thereby bypassing VAE compression and enabling end-to-end optimization directly in RGB space. Complementarily, the latter decomposes the asymmetric flow target into a low-rank semantic component and a noise-free orthogonal texture component. By routing these components to specialized branches, it achieves decoupled optimization with reduced mutual interference. Benefiting from these designs, our PixelSR restores perceptually realistic details while faithfully preserving structural content. Extensive experiments across multiple benchmarks demonstrate that PixelSR not only achieves competitive performance, but also effectively mitigates the fidelity--realism seesaw effect.

cs.CV↗

arXiv2026

Dynamic Constraint Reconstruction Based Control Barrier Functions for Safety-Critical Control of High-Dimensional Manipulators

Control barrier functions (CBFs) provide formal safety guarantees for constrained nonlinear systems, but their effectiveness relies on accurate system dynamics. In high-dimensional manipulators subject to unknown disturbances and model uncertainties, fixed safety constraints constructed from nominal dynamics may become inconsistent with the actual system behavior, leading to safety degradation or excessive conservatism. This paper proposes a dynamic constraint reconstruction based control barrier function (DCR-CBF) framework for safety-critical control of disturbed robotic manipulators. An extended state observer is employed to estimate lumped disturbances online, and the estimated disturbance is incorporated into high-order control barrier functions to reconstruct safety constraints according to the estimated true dynamics. To address estimation inaccuracies, a safety margin is introduced, and a sufficient condition is derived to guarantee forward invariance under bounded estimation errors. Simulation studies on a 4-DOF excavation manipulator demonstrate that the proposed DCR-CBF method achieves zero safety violation under strong unknown disturbances while significantly improving trajectory-tracking performance compared with standard and robust CBF methods.

eess.SY↗

arXiv2026

Dirac branch-cut modes with relativistic transport

Emergent Dirac fields, exhibiting effective relativistic physics, are most commonly associated with bulk and surface states in materials such as graphene and topological insulators. Here we show that phase dislocations of a complex Dirac mass support a distinct family of propagating Dirac defect states that behave as massless or massive one-dimensional relativistic modes. These traveling-wave defect states, termed Dirac branch-cut (DBC) modes, are governed by a one-dimensional Dirac Hamiltonian whose effective mass is controlled by the phase difference across the branch cut. Using acoustic metamaterials, we experimentally demonstrate a range of relativistic phenomena exhibited by DBC modes, including relativistic dispersion, energy-independent confinement, Klein tunnelling, and transport along freeform (e.g., spiral) trajectories. Our results establish branch-cut defects as a distinct platform for realizing propagating Dirac defect states, extending relativistic Dirac physics from bulk and surface states to propagating modes confined to defect boundaries.

physics.optics↗

arXiv2026

Heterogeneous Tactile Transformer

Tactile sensors are inherently heterogeneous: a model trained on one sensor cannot be directly used on another, which limits learning contact-rich manipulation policies from diverse tactile data at scale. To bridge this gap, we propose the Heterogeneous Tactile Transformer (HTT), a framework that learns shared tactile representations across heterogeneous sensors. HTT consists of sensor-specific encoders and a shared transformer trunk, and is pretrained with per-modality masked reconstruction together with cross-modal alignment between paired sensors. Pretraining uses our novel Heterogeneous Paired Tactile (HPT) dataset, containing 1.6M synchronized paired frames across four vision- and array-based tactile sensors. Across distinct tactile perception and real-world manipulation tasks, HTT is shown to learn transferable representations that adapt to new tasks and previously unseen sensors. Dataset, code, and model checkpoints are released at https://jxbi1010.github.io/htt-gh-page/.

cs.RO↗

arXiv2026

Fewer Tokens, Better Action: GPT-6 Astra Robot Agents with 14% Higher Success Rate but 65% Fewer Tokens

Vision-language model (VLM) agents can control robots through visual feedback and action primitives, but repeated LLM calls and redundant observations incur substantial token overhead. We introduce PyRUA-Lean, an interactive code-execution framework in which the agent composes robot primitives, including vision-language-action (VLA) policies, into Python cells that check and retry locally and return only the images and state they request. Across 700 simulated task instances from LIBERO-PRO, RoboTwin 2.0, and RoboCasa365, it raises success from 63.1% to 71.7% over tool calling with the same GPT-6 Astra planner and primitives, using 49% fewer LLM calls and 65% fewer input tokens on jointly solved instances. On RoboDojo, without any VLA policy or privileged information, it reaches 51.6% average success rate (Score 58.2), 19.4 points above the top entry of the public leaderboard. Based on the PyRUA-Lean framework, we train PyRUA-Lean Coder v0.1-4B, an open 4B model, with GRPO on 45 LIBERO-PRO instances; it improves from 18.8% to 44.1% on unseen task instances, surpassing the Qwen3.5-27B base model.

cs.CV↗

arXiv2026

Accelerating MoE Training through Expert Placement and Replica Coordination

Distributed training of sparsely activated mixture-of-experts (MoE) models is often bottlenecked by workload imbalance and all-to-all (A2A) communication induced by dynamic token routing. Existing token deduplication, expert placement, and replication techniques alleviate these bottlenecks but do not jointly capture their distinct load semantics. Under hierarchical token deduplication, communication counts each token once per destination group at each hierarchy level, whereas expert computation processes every token-expert assignment. With replication, both costs depend jointly on physical layout and token-to-copy mapping, making separate optimization suboptimal. We present PlaceMoE to address this coupling. First, we formulate the joint optimization of expert layout and source-aware token-to-copy mapping for E experts distributed across P devices, accounting for both hierarchical communication traffic and expert-compute load. Second, we develop a solver that combines co-selection-aware expert grouping, bounded dynamic programming for replica allocation, and alternating refinement of layout and mapping. Third, PlaceMoE overlaps replica-gradient synchronization with attention backward. We evaluate PlaceMoE on a 32-GPU cluster and a 64-NPU cluster using Qwen3-VL and DeepSeek-V3. Experimental results show that PlaceMoE achieves up to a 6.94x speedup in A2A communication and 1.74-2.33x speedups in end-to-end training over the state-of-the-art training framework VeOmni.

cs.DC↗

arXiv2026

Macroscopic study of $ϕ$-mesic nuclei in $pp$ collisions at $\sqrt{s}$ = 7 TeV with the PACIAE model

A macroscopic method of the parton and hadron cascade model PACIAE combined with the coalescence model DCPC is employed to study the $ϕ$-mesic nuclei in $pp$ collisions at $\sqrt{s}$ = 7 TeV for the first time in this work. Based on the final hadronic states of $pp$ collisions simulated by the PACIAE model, the $ϕ$-mesic nuclei: $\mathrm{^3_ϕ H}(ϕ-pn)$, $\mathrm{^4_ϕ H}(ϕ-pnn)$, and $\mathrm{^4_ϕ He}(ϕ-ppn)$ are recombined by the DCPC model with constituents of protons, neutrons, and $ϕ$ mesons. Their yields, transverse momentum spectra, and rapidity distributions are simulated separately. These theoretical predictions call for experimental confirmation.

nucl-th↗

arXiv2026

Factorization-Guided Training and Inference for Diffusion-Drafted Speculative Decoding

Speculative decoding accelerates large language model inference through parallel verification of tokens proposed by a drafter. Diffusion drafters further parallelize proposal generation, but greater proposal capacity or higher mean accepted tokens (MAT) need not yield higher throughput. Because verification accepts only a consecutive prefix, early rejections can render later correct predictions ineffective, limiting MAT gains, while even higher MAT may not improve throughput if per-round latency increases. We introduce DScout, which uses two complementary factorizations to guide diffusion drafter training and inference for higher throughput. For training, DScout factorizes position-wise loss into inclusion, weighting, and prediction loss. This guides the design and combination of mixed-horizon training, intra-block unmasking, and acceptance-weighted loss to align supervision with prefix acceptance and improve MAT. For inference, DScout factorizes throughput as the ratio of MAT to per-round latency, guiding target-conditioning depth, drafter depth, and inference block size selection to identify a high-throughput configuration for each workload. Across six benchmarks, DScout achieves $3.84\times$ and $4.14\times$ mean speedups over autoregressive decoding on Qwen3-4B and Qwen3-8B, respectively, improving mean speedup over the state-of-the-art baseline by 9.1% and 7.1%. Across models and concurrency settings, our inference configuration policy provides up to 9.75% additional mean speedup over DScout's fixed configuration.

cs.LG↗

arXiv2026

Physics-Informed Conditional Diffusion for Motion-Robust Retinal Temporal Laser Speckle Contrast Imaging

Retinal laser speckle contrast imaging (LSCI) is a noninvasive optical modality for monitoring retinal blood flow dynamics. However, conventional temporal LSCI (tLSCI) reconstruction relies on sufficiently long speckle sequences to obtain stable temporal statistics, which makes it vulnerable to acquisition disturbances and limits effective temporal resolution. A physically informed reconstruction framework, termed RetinaDiff (Retinal Diffusion Model), is proposed for retinal tLSCI that is robust to motion and requires only a few frames. In RetinaDiff, registration based on phase correlation is first applied to stabilize the raw speckle sequence before contrast computation, reducing interframe misalignment so that fluctuations at each pixel primarily reflect true flow dynamics. From the long registered sequence this step yields a high-quality multiframe tLSCI map that serves only as the reconstruction target, while a motion-corrected contrast prior is computed independently from the few input frames. Next, guided by this prior, a conditional diffusion model performs inverse reconstruction by jointly conditioning on the registered few-frame sequence and the prior. On stable sequences acquired with an in-house retinal LSCI system, RetinaDiff improved SSIM from 0.159 to 0.533, PSNR from 14.83 to 18.06 dB, and FID from 211.50 to 111.55 compared with direct five-frame reconstruction, showing improved structural continuity and statistical stability over representative baselines. The framework also remains effective in a small number of extremely challenging cases, where both the direct five-frame input and the conventional multiframe reconstruction are severely degraded. Overall, this work provides a practical and physically grounded route for reliable retinal tLSCI reconstruction from extremely limited frames. The source code and model weights will be released upon acceptance.

cs.CV↗

arXiv2026

Large class of many-to-one mappings over quadratic extension of finite fields

Many-to-one mappings (including $1$-to-$1$ mappings) over finite fields play important roles in cryptography and coding theory. This paper provides a unified framework for studying the many-to-one property of polynomials of the form $f(x) = h(a x^q + b x + c) + u x^q + v x$, where $h(x) \in \mathbb{F}_{q^2}[x]$ and $a$, $b$, $c$, $u$, $v \in \mathbb{F}_{q^2}$. Our approach uses two linear transformations from $\mathbb{F}_{q^2}$ to $\mathbb{F}_{q}$ to establish a commutative diagram relating $f(x)$ to an associated polynomial $g(x)$. This diagram reduces the characterization of the many-to-one property of $f(x)$ on $\mathbb{F}_{q^2}$ to that of $g(x)$ on the subfield $\mathbb{F}_{q}$. In particular, when $h(x) = x^{r}$ and~$r$ satisfies suitable conditions, we show that $g(x)$ can be reduced to either a low-degree polynomial or a linearized binomial. Then, employing the known many-to-one properties of these low-degree or linearized polynomials over~$\mathbb{F}_{q}$, we derive explicit conditions that completely characterize when $f(x)$ is many-to-one on $\mathbb{F}_{q^2}$. Moreover, we determine the inverses of all $1$-to-$1$ mappings obtained in this paper, and we show that the $2$-to-$1$ mappings of this form naturally yield explicit involutions. Our findings generalize and unify many results in the literature.

cs.IT↗

arXiv2026

Optical Möbius Snails

The exploration of topological states has emerged as a vibrant frontier across diverse physical systems. Among these, the Möbius strip stands out as a canonical structure that has recently inspired intriguing optical physics.In this work, we extend this concept into the high-dimensional realm by using an optical analogue-the optical Möbius snail. Specifically, employing a skyrmionic beam as the input field, tight focusing yields a continuous family of Möbius strips that exists in three-dimensional space, giving rise to a snail-like polarization topological texture. Furthermore, we demonstrate full-field polarization tomography and achieve switchable control of this optical snail texture by tuning the Skyrme number of the incident skyrmionic beam. Our findings establish a direct bridge between low- and high-dimensional Möbius geometries, opening a new platform for exploring complex spatial topology and tailored light-matter interactions.

physics.optics↗

arXiv2026

Accelerating Disaggregated RL for Visual Generative LLMs with Diffusion-Based Parallelism and Trainer-Assisted Generation

Reinforcement learning (RL) has become a dominant post-training paradigm, driving the emergence of high-performance RL systems such as veRL for autoregressive large language models (LLMs). In parallel, diffusion-oriented RL algorithms, e.g., DanceGRPO and FlowGRPO, have rapidly expanded the scope of RL from language reasoning to diffusion-based visual and flow-based generation. However, efficient RL systems for diffusion generative LLMs remain underexplored. Existing implementations, e.g., veRL-Omni, still rely on colocated execution, which simplifies synchronization but couples rollout and training resources, limits heterogeneous deployment, and constrains independent scaling. To this end, we introduce DigenRL, a disaggregated RL framework for diffusion-based generative LLMs that supports flexible resource allocation, accommodates heterogeneous GPUs, and facilitates efficient task scheduling. To maximally reduce the execution bubbles in the disaggregated architecture, we propose: 1) a generation-axis pipeline (GAP) and time-step parallelism (TSP) in the diffusion architecture to enable finer-grained pipelining between rollout and training; 2) an elastic trainer-assisted generation (TAG) approach to enable the trainer GPU resources to dynamically assist in executing rollout generations; and 3) a tightly one-step constrained asynchronous strategy to further utilize the tail bubble in the pipeline. Extensive experiments are conducted on three hardware testbeds with 16-32 GPUs using HunyuanVideo-13B, Wan2.1-14B, FLUX.1-12B, and QwenImage-20B generative models. Experimental results show that DigenRL achieves 1.56-2.10x throughput improvements over state-of-the-art diffusion RL systems, veRL-Omni and GenRL.

cs.AI↗

arXiv2026

A generalization of the map $χ$

The mapping $ χ_n:\mathbb{F}_2^n \to \mathbb{F}_2^n$ defined by $y=χ_n(x)$ with $y_i = x_i + x_{i+1}x_{i+2} + x_{i+2}$, where the indices are computed modulo $n$, has been widely studied for its application in lightweight cryptography. In this paper, we generalize this mapping and completely characterize all these shift-invariant permutations of the form $y_i=x_{i+u}+x_{i+v}(x_{i+w}+a_i)$, where $0\le u, v, w<n$ and $a_i\in \mathbb{F}_2$, $1\le i\le n$.

cs.IT↗

arXiv2026

How to Better Train VLAs: Lessons Learned From the REAL-I Challenge at ICRA 2026

How can robot policies learn more effectively from a fixed demonstration budget? The first Real-world Embodied AI Learning (REAL-I) Challenge at ICRA 2026 examined this question through simulation, real-robot evaluation, and an on-site final on a shared dual-arm humanoid platform. We describe the challenge tasks, data and deployment interfaces, and competition results, then compare the approaches contributed by NUS-CLEAR, RCL-Lab, and DeepTouch AI. Their systems combined pretrained vision-language-action models and task-specific imitation policies with different strategies for data curation, staged adaptation, checkpoint selection, and action-space design. The team reports highlight the importance of adapting to the deployment environment while retaining prior capabilities, treating demonstration quality at an appropriate temporal scale, and suppressing errors in inactive robot components. They also expose the limitations of offline action-prediction metrics for forecasting closed-loop success. These observations motivate a view of fixed-data robot learning that integrates data, adaptation, evaluation, and deployment.

cs.RO↗