arXiv subjects

Pengfei Wang

Publications and source records attributed to Pengfei Wang.

At least 19 recordsLinked to original sources

arXiv2026

Finite-Sample Calibration and Power Benchmarks in Two-Sample Random-Integration Tests

In high-dimensional two-group studies, a supplied variable set can be used directly or to learn a projection. We study how much attainable power these choices retain at fixed observation and calibration budgets. For complete unbiased random-integration statistics under Gaussian common shape, a fixed-spectrum comparison orders direction-wise power and gives an explicit criterion balancing an oracle advantage against learning loss. Joint regeneration controls level for a specified learner with unknown diagonal shape. A paired construction permits observation reuse and finite-reference calibration of fitted scores and menus. In the paired diagonal experiment, an invariant sparse likelihood-ratio benchmark provides an attainable efficiency target. Finite-amplitude bounds connect a classical bounded quadratic to this target; conditioning on the coordinate maximum yields informative complementary bounds at moderate and strong signals. Both comparisons include finite reference budgets, and null-only simulations certify local comparisons with specified fitted rules. Equal-budget experiments and a leukemia application compare aggregation, projection and fitted quadratic scores. For correlated paired data, classical sign randomization supplies a separate calibration under central symmetry. The efficiency guarantees apply to the stated experiments and signal classes; general small-budget shape efficiency and adaptation to unknown signal allocation remain open.

stat.ME↗

arXiv2026

Beyond Semantic IDs: Atomic Creator Tokens for Large-Scale Generative Recommendation

Recommending creators that a user does not yet follow is a cold-start problem: from hundreds of thousands of popular creators and only a short trail of prior engagements, the system must predict the next creator worth surfacing. The prevailing generative recommendation recipe tokenizes each creator into a \emph{Semantic ID} (SID) by quantizing a frozen embedding and trains a language model to generate SIDs. We show that this SID layer, not the language model, is the binding constraint. The creator-to-SID map is many-to-one, so colliding creators must be resolved by a popularity tie-break that caps creator-level recall and favors already-popular creators, and the frozen quantizer stops the training gradient from reaching the creator representation. We propose Atomic Creator Token (ACT), which gives every creator its own vocabulary token trained end-to-end. The map is bijective, so collisions are impossible and inference is a direct top-$k$ over creator logits. In a controlled comparison on production data in which the token layer is the only variable, ACT improves creator-level Recall@15 by 19.8 points ($1.84\times$) over a matched SID baseline, still leads when the SID codebook is enlarged $64\times$, and serves at ${\sim}2.7\times$ lower per-user latency. The advantage carries over to the public KuaiRand dataset, where ACT is the only method that beats a global-popularity baseline, provided each creator token sees enough training examples. In a live A/B test, ACT is directionally positive on every reported metric against both the SID-enhanced and the classic production rankers.

cs.AI↗

arXiv2026

S3KF: Spherical State-Space Kalman Filtering for Panoramic 3D Multi-Object Tracking

Panoramic multi-object tracking is important for industrial safety monitoring, wide-area robotic perception, and infrastructure-light deployment in large workspaces. In these settings, the sensing system must provide full-surround coverage, metric geometric cues, and stable target association under wide field-of-view distortion and occlusion. Existing image-plane trackers are tightly coupled to the camera projection and become unreliable in panoramic imagery, while conventional Euclidean 3D formulations introduce redundant directional parameters and do not naturally unify angular, scale, and depth estimation. In this paper, we present $\mathbf{S^3KF}$, a panoramic 3D multi-object tracking framework built on a motorized rotating LiDAR and a quad-fisheye camera rig. The key idea is a geometry-consistent state representation on the unit sphere $\mathbb{S}^2$, where object bearing is modeled by a two-degree-of-freedom tangent-plane parameterization and jointly estimated with box scale and depth dynamics. Based on this state, we derive an extended spherical Kalman filtering pipeline that fuses panoramic camera detections with LiDAR depth observations for multimodal tracking. We further establish a map-based ground-truth generation pipeline using wearable localization devices registered to a shared global LiDAR map, enabling quantitative evaluation without motion-capture infrastructure. Experiments on self-collected real-world sequences show decimeter-level planar tracking accuracy, improved identity continuity over a 2D panoramic baseline in dynamic scenes, and real-time onboard operation on a Jetson AGX Orin platform. These results indicate that the proposed framework is a practical solution for panoramic perception and industrial-scale multi-object tracking.The project page can be found at https://kafeiyin00.github.io/S3KF/.

cs.RO↗

arXiv2026

Memorize When Needed: Decoupled Memory Control for Spatially Consistent Long-Horizon Video Generation

Spatially consistent long-horizon video generation aims to maintain temporal and spatial consistency along predefined camera trajectories. Existing methods mostly entangle memory modeling with video generation, leading to inconsistent content during scene revisits and diminished generative capacity when exploring novel regions, even when trained on extensive annotated data. To address these limitations, we propose a decoupled framework that separates memory conditioning from generation. Our approach significantly reduces training costs while simultaneously enhancing spatial consistency and preserving the generative capacity for novel scene exploration. Specifically, we employ a lightweight, independent memory branch to learn precise spatial consistency from historical observations. We first introduce a hybrid memory representation to capture complementary temporal and spatial cues from generated frames, then leverage a per-frame cross-attention mechanism to ensure each frame is conditioned exclusively on the most spatially relevant historical information, which is injected into the generative model to ensure spatial consistency. When generating new scenes, a camera-aware gating mechanism is proposed to mediate the interaction between memory and generation modules, enabling memory conditioning only when meaningful historical references exist. Compared with the existing method, our method is highly data-efficient, yet the experiments demonstrate that our approach achieves state-of-the-art performance in terms of both visual quality and spatial consistency. Project page can be found at https://iguoyanjun.github.io/memorize-when-needed-project-page/.

cs.CV↗

arXiv2026

OSWAM: An Effective World Action Model with One-Step Action Generation

Direct-policy World Action Models (WAM) have demonstrated the benefits of decoupling action generation from explicit future visual prediction for efficient inference in autonomous driving and embodied control. However, they still require tens of denoising steps to achieve competitive action quality, incurring substantial latency and hindering real-time deployment. Simply reducing the number of denoising steps will collapse the nonlinear multi-step denoising trajectory into coarse updates, substantially degrading action quality. To reduce the denoising steps while enabling real-time action generation, we present an effective world action model with one-step action generation, namely OSWAM, for autonomous driving and embodied control. We argue that it is the absence of structured action priors in random noise initialization that forces direct-policy WAMs to recover coherent action sequences through iterative denoising. Hence, we construct imagination tokens to initialize the diffusion process, providing physical priors and action-conditioned representations. We further introduce privileged information distillation with future conditioning to enhance future-aware action generation. As a result, OSWAM can efficiently and effectively generate high quality actions in just one diffusion step. Our experiments demonstrate that OSWAM achieves comparable or even better results in both autonomous driving and embodied control than direct-policy WAM and Joint-WAM that require tens of steps. It obtains approximately $3\times$ speedup over direct-policy WAM baselines and supports real-time compiled inference with latency as low as 40ms on an NVIDIA RTX 5090 GPU. The project page is available at https://gwenzhang.github.io/OSWAM/

cs.CV↗

arXiv2026

Zeeman Shift and Injection Currents under Light Irradiation

The conversion of light into electrical currents enables energy harvesting and photodetection. Conventional second-order dc photocurrents are driven by the electric component of light through electric-dipole coupling and require broken inversion symmetry. Here, we show that the electric and magnetic components of light jointly generate Zeeman shift and injection currents. Using quantum perturbation theory, we derive explicit microscopic expressions for their conductivity tensors. We systematically classify the symmetry constraints on these responses and show how they differ from those governing conventional photocurrents. In particular, both responses can occur in centrosymmetric systems with spin-orbit coupling or noncollinear magnetic order, irrespective of time-reversal symmetry. A centrosymmetric kagome model exhibits broadband responses. First-principles calculations for Mn$_3$Sn further predict Zeeman injection and shift current densities of $10^{2}$-$10^{3}~\mathrm{A/m^{2}}$ and $1$-$10~\mathrm{A/m^{2}}$, respectively, under illumination with an electric-field amplitude $E=10^{4}~\mathrm{V/m}$. These magnitudes place both currents within experimental reach.

cond-mat.other↗

arXiv2026

An F-Type Normal-Reference Calibration for High-Dimensional Covariance Testing

An F-type normal-reference calibration is developed for testing equality of two high-dimensional covariance matrices with unknown means. The test statistic is a ratio whose numerator measures the Frobenius covariance discrepancy and whose denominator estimates its null expectation. Their difference is the complete Li-Chen statistic. Under suitable moment and trace conditions, we establish a Gaussian-reference approximation. After standardization, the statistic and its reference share the normal or nonnormal limits determined by a convergent normalized spectrum. The reference is a ratio of independent chi-square-type mixtures. Applying Welch-Satterthwaite matching to its two components gives an ordinary F approximation, with degrees of freedom estimated from independent four-observation blocks. The effect of estimating the population means is controlled by an exact expansion, and local-power and consistency results are established. The ordinary F approximation is asymptotically valid under a diffuse spectrum, whereas a covariance spike can produce a nonvanishing approximation error. Simulation studies examine empirical size and power under normal and nonnormal sampling. Averaging the trace estimates over fixed partitions improves their stability, but does not remove size inflation in the strongly correlated settings.

math.ST↗

arXiv2026

Continuous-wave 148.4-nm generation in strontium tetraborate for nuclear absorption spectroscopy

Continuous-wave light near 148.4 nm enables direct absorption spectroscopy of the thorium-229 nuclear transition. We demonstrate single-pass frequency doubling in strontium tetraborate, delivering an estimated 8.1 nW at 920 mW pump power. At this power, oxygen admission leaves a background of 1.22 kHz, corresponding to an estimated signal-to-background ratio of 3.8 x 10^4 after current-integration calibration. Several crystals grown by top-seeded solution growth produce polarization-dependent VUV signals. Temperature scans reveal distinct maxima among specimens cut from one boule and among positions within a specimen. The source sustains operation over 33.3 h. In a separate 3-h record acquired with 1-ms gates, local referencing achieves near-Poisson averaging, reaching a fractional Allan deviation of 3.5 x 10^-5 at 1000 s. These results establish a practical VUV source for stable, long-duration spectroscopy of weak nuclear absorption.

physics.optics↗

arXiv2026

A solid-state nuclear clock based on VUV absorption spectroscopy of $^{229}$Th

We demonstrate sustained operation of a solid-state thorium-229 nuclear clock using continuous-wave absorption at 148.4 nm. In a $^{229}\mathrm{Th}:\mathrm{CaF}_2$ crystal, we resolve four quadrupole components of the D center and a broader O-center resonance. Feedback on the strongest narrow component is updated approximately every 10 s and remains active throughout a 30-h record. A hydrogen-maser-referenced frequency comb measures the output. Its fractional frequency instability follows approximately $1.29\times10^{-11}(τ/\mathrm{s})^{-1/2}$ and reaches $1.24\times10^{-13}$ at $10^4$ s. A separate 6.6-h run demonstrates feedback on a second quadrupole component. These measurements connect the resolved absorption spectra with sustained nuclear-clock operation and frequency comparisons between crystal segments.

physics.atom-ph↗

arXiv2026

Decoding ML Decision: An Agentic Reasoning Framework for Large-Scale Ranking System

Modern large-scale ranking systems operate within a sophisticated landscape of competing objectives, operational constraints, and evolving product requirements. Progress in this domain is increasingly bottlenecked by the engineering context constraint: the arduous process of translating ambiguous product intent into reasonable, executable, verifiable hypotheses, rather than by modeling techniques alone. We present GEARS (Generative Engine for Agentic Ranking Systems), a framework that reframes ranking optimization as an autonomous discovery process within a programmable experimentation environment. Rather than treating optimization as static model selection, GEARS leverages Specialized Agent Skills to encapsulate ranking expert knowledge into reusable reasoning capabilities, enabling operators to steer systems via high-level intent vibe personalization. Furthermore, to ensure production reliability, the framework incorporates validation hooks to enforce statistical robustness and filter out brittle policies that overfit short-term signals. Experimental validation across diverse product surfaces demonstrates that GEARS consistently identifies superior, near-Pareto-efficient policies by synergizing algorithmic signals with deep ranking context while maintaining rigorous deployment stability.

cs.AI↗

arXiv2026

OREO: Fidelity Alignment in 3D Generation via On-the-fly Rendering-Editing Optimization

Despite recent advancements in 3D generation, models often struggle to produce assets with high visual fidelity. To bridge this gap, we propose OREO, an alignment framework that enhances the realism of 3D generators by leveraging rich 2D diffusion priors. Instead of relying on static datasets, OREO establishes a dynamic optimization loop that produces on-the-fly edited renderings as 2D pseudo-targets. At its core, we introduce Reinforced Editing, which utilizes a 2D model to refine rendered views of the 3D output, enhancing their overall visual fidelity while preserving the underlying geometry, viewpoint, and content. These refined views serve as high-quality supervision targets, enabling the 3D generator to learn from its own generated samples and progressively improve its visual quality. Experiments demonstrate that OREO effectively improves upon pre-trained baselines, producing 3D assets with enhanced visual realism.

cs.CV↗

arXiv2026

Bounding Selmer Groups of Superelliptic Jacobians via Class Groups

Let $K$ be a number field containing a primitive $p$-th root of unity $ζ_p$. Let $f(x)\in K[x]$ be a monic integral polynomial, and let $f_0$ denote its radical. Let $C/K$ be the superelliptic curve defined by $y^p=f(x)$, and let $J$ be its Jacobian variety. The variety $J$ admits multiplication by $ζ_p$ over $K$; in particular, the endomorphism induced by $Π:=1-ζ_p$ gives an isogeny of $J$ over $K$. Let $\operatorname{Sel}_Π(J)$ denote the Selmer group associated to $Π$. Under suitable hypotheses, we obtain bounds for $\operatorname{Sel}_Π(J)$ in terms of the $p$-torsion subgroup of the class group of $L:=K[x]/(f_0)$. Several examples illustrating the results are also discussed.

math.NT↗

arXiv2026

PrimSynth: An Agentic Approach to Discover, Validate, and Synthesize Exploit Primitives for Linux Kernel Vulnerabilities

Linux kernel vulnerabilities are critical to downstream systems. Despite extensive research on automated kernel exploitation, a fundamental challenge remains the conceptual gap between abstract exploit strategies and concrete technical operations. To fill this gap, this paper introduces a systematic characterization that formalizes six classes of exploit primitives from logical capability to validatable effect. Then, an extended exploit strategy representation is proposed, which couples primitive upgrading strategies with primitive path code synthesis rules governing object constraints, temporal sequencing, environment prerequisites, and validation constraints. Building upon this foundation, this paper presents \textsc{PrimSynth}, a multi-agent framework that encapsulates these representations through coordinated agents to discover, validate, and synthesize exploit primitives for memory corruption vulnerabilities in the Linux kernel. These agents operate in an iterative closed loop until valid primitives are found, leveraging validation signals as evidence of exploitable state transitions to ground primitive synthesis decisions. An automated method for extracting and validating primitives is also proposed based on vulnerability-directed execution and a rebootable validation environment. \textsc{PrimSynth} is evaluated on 16 real-world Linux kernel CVEs spanning 5 vulnerability types. Experimental results show that PrimSynth achieves reliable primitive extraction, maintaining a 100% primitive match rate. For primitive synthesis, PrimSynth successfully synthesizes multi-primitive exploitation chains with 82.4% strategy synthesis rate (SSR) when the public PoC is available and a 61.3% SSR without the guidance of primitive hypotheses.

cs.CR↗

arXiv2026

A compact vapor-cell optical frequency reference with fractional frequency instability around $10^{-16}$

Compact optical frequency reference with high stability is essential for field applications such as navigation and geodesy, yet vapor cell systems have remained confined to fractional instabilities over $10^{-15}$. Here, we report a molecular iodine reference that reaches an instability of $7 \times 10^{-16}$ at 1000 s and operates at the $10^{-16}$ level from 200 to 2000 s, surpassing the best reported vapor cell standards by approximately a factor of three. This achievement is enabled by a monolithic, drift immune spectroscopic unit bonded to an ultra low expansion glass substrate with precision control of key parameters.The entire system occupies only 25 L.The system achieves $5 \times 10^{-15}$ instability at 1 s and reaches the $10^{-16}$ level over the 200 to 2000 s averaging-time range, representing the first medium term stability at the $10^{-16}$ level from a compact, field ready vapor-cell reference. Our work demonstrates that $10^{-16}$ instability can be engineered into portable systems, opening a path to high precision time-keeping beyond the laboratory.

physics.atom-ph↗

arXiv2026

Fixed-time stabilization of systems with nilpotent matrices in the presence of large unknown delays

In this paper, we study state-feedback fixed-time stabilization of continuous-time perturbed linear systems with nilpotent matrices in the presence of unknown large constant input/measurement delays with known bounds. By using an additional artificial delay, we design a feedback that for a precisely known linear system with a nilpotent matrix drives the state to zero in fixed time (which is larger than the bound of the unknown delay) and achieves input-to-state stability with a large exponential decay rate in the presence of additive disturbances and small system matrix perturbations and nonlinearities. For perturbed single integrators, fixed-time stabilization is achieved in the case of fast-varying unknown measurement delays with known bounds. The results are extended to discrete-time systems. Numerical simulations illustrate the efficiency of the results.

math.OC↗

arXiv2026

Room-temperature ferroelectrically switchable quantum geometry in few-layer WTe2 for complementary in-memory computing

Quantum geometry, describing the inherent geometric structure of electron wavefunctions in momentum space, transcends the traditional charge degree of freedom and provides a novel physical basis for information encoding and processing. The key to such new computing paradigms is the non-volatile electrical programming of quantum geometric states at room temperature, which, however, has not been established. Here, we demonstrate ferroelectrically switchable quantum geometry in few-layer WTe2, which uniquely enables complementary convolutional processing. By employing the intrinsic coupling between ferroelectric polarization and quantum geometry in few-layer WTe2, we show that the second- and third-order nonlinear anomalous Hall effects (NLAHE) can be deterministically and electrically switched in a nonvolatile and correlated manner. The switching is robust at room temperature for ~104 cycles and retention of ~105 s. Furthermore, leveraging the opposite switching behaviors of second- and third-order NLAHE at room temperature, we demonstrate complementary in-memory computing and implement a hardware-level complementary convolution kernel. This kernel overcomes the inherent directional specificity of conventional convolutional networks and achieves a texture recognition accuracy of 98%, thereby illustrating a viable pathway towards physics-native computing through exploiting exotic physics in quantum materials.

cond-mat.mtrl-sci↗

arXiv2026

MOSS-VL Technical Report

We present MOSS-VL, an open vision-language model family that treats real-time interaction -- perceiving while it speaks -- as a first-class capability. It is co-designed across the stack: the language decoder attends to vision only through gated cross-attention, so the model can naturally see incoming frames while generating; a synthesized interaction corpus supervises when to speak, when to stay silent, and when to revise; and a staged curriculum concentrates all real-time-specific training in one light final stage over a strong offline foundation. Offline, MOSS-VL-Instruct is competitive at comparable scale and leads temporal-reasoning video sets. Across four streaming benchmarks, MOSS-VL-Realtime posts the best average on three (second on the fourth) among open-source streaming models, sweeping the three subsets that squarely test proactive behavior -- 66.0 vs. 37.5 for the best baseline on OmniMMI Proactive Alerting. With 11.3B parameters but visual tokens outside the decoded sequence, MOSS-VL widens its time-to-first-token advantage over same-backbone Qwen3-VL-8B from 2.8x to 5.1x as visual context grows. We release all five checkpoints, the training curriculum, and the real-time inference code at https://github.com/OpenMOSS/MOSS-VL.

cs.CV↗

arXiv2026

Salience Induction against Multi-Hop RAG Agents: Threat and Defense

Agentic retrieval-augmented generation (RAG) systems increasingly retrieve external evidence and orchestrate tools for knowledge-intensive applications. In Multi-Hop question answering, agents chain facts across documents. Existing defenses focus on content poisoning, which injects false facts, and prompt injection, which embeds directives. We identify a third attack surface: the salience channel, through which fact position, emphasis, framing, and semantic proximity can redirect reasoning even when all retrieved claims are true and no instructions are present. We formalize Salience Induction as truth-preserving edits that redirect Multi-Hop attribute binding while leaving the retrieval trace semantically intact. We define six Salience-Editing operator classes and build an iterative proposer-verifier pipeline under factual and stealth constraints. We also introduce SalientWiki-MH, a decoy-annotated Multi-Hop benchmark. Evaluations across five frontier model families (GPT, Claude, Gemini, DeepSeek, and Qwen) and three agent architectures (ReAct, Reflexion, and tool-calling) show broad generalization. Under a 30% edit budget, Salience Induction achieves an 83.3% attack success rate; the strongest evaluated baseline defense leaves 75.7% post-defense ASR. Untargeted rewriting further reduces attacks only by degrading neutral task success. Our lightweight input-side defense, Salience Normalization, reduces attack success to 15.3% under standard attacks and 23.6% under an adaptive attack. These results show that truthfulness and instruction filtering alone are insufficient: robust agentic RAG also requires defenses against salience-relevance decoupling.

cs.CR↗