arXiv subjects

Wenyi Zhang

Publications and source records attributed to Wenyi Zhang.

At least 19 recordsLinked to original sources

arXiv2026

VARG: Value-Aware and Ranking-Aligned Generative Retrieval for Dynamic E-commerce Search

Integrating recall and pre-ranking in e-commerce search requires candidate generation to account for relevance, personalization, and business value before final ranking. To this end, we present VARG, a generative retrieval system for Tmall App search that directly admits generated item candidates to the existing final ranker. VARG-ID constructs semantic prefixes using RQ-VAE, enhances search relevance through bidirectional query-item contrastive learning, and combines these prefixes with a value-ordered third token to provide fine-grained item addresses and a business-value prior. Three-stage supervised fine-tuning progressively learns item-to-identifier mappings, query-semantic retrieval, and personalized retrieval. Personalized model training combines value-aware and hierarchy-aligned supervision with expanded user context, and uses local ordinal supervision (LO-SFT) to learn the local within-cluster ordering encoded by the third token. Prefix-GRPO combines gated rewards based on output legality, user behavior, ranker advantage, and search relevance with prefix-aware token weighting to align candidate generation with business value and ranking objectives. Coordinated daily product and model updates preserve existing item addresses while incorporating new products and behavioral feedback. Offline experiments on tens of millions of products validate identifier stability and demonstrate gains in retrieval quality and head-level value recall from SFT strategies and Prefix-GRPO over their respective baselines. In a 14-day online A/B test covering 20% of search traffic, VARG directly admits generated candidates to the final ranker and improves GMV by 1.45%, per-user IPV by 0.22%, and PCTR by 0.31%. Online shopping-guide query evaluations further show that VARG maintains competitive relevance with a smaller candidate quota.

cs.IR↗

arXiv2026

A Finite-Blocklength Analysis for ORBGRAND

Within the Guessing Random Additive Noise Decoding (GRAND) family, ordered reliability bits GRAND (ORBGRAND) offers hardware-friendly use of soft information. Existing information-theoretic results for ORBGRAND are asymptotic in blocklength and do not quantify its short-to-moderate blocklength performance. This paper develops a finite-blocklength analysis for ORBGRAND over general binary-input memoryless channels, addressing the key challenge that the rank-induced decoding metrics are coupled across symbols. We specialize the random-coding union (RCU) bound for mismatched decoding to ORBGRAND, yielding the ORB-RCU bound. We then characterize the two codeword decoding metrics governing this bound: the transmitted-codeword decoding metric is represented, up to a deterministic $O(n^{-1})$ term, by a nondegenerate second-order $U$-statistic and admits a Gaussian approximation via the Hoeffding projection and a Berry-Esseen theorem for $U$-statistics, whereas the competing-codeword decoding metric is reduced to a weighted sum of independent and identically distributed Bernoulli random variables and characterized by a strong large-deviation expansion that holds uniformly. Combining these ingredients, we establish a Gaussian approximation for the ORB-RCU bound, which yields a third-order achievable rate expansion for ORBGRAND. The first-order term coincides with the previously established ORBGRAND achievable rate, the second-order term defines an ORBGRAND dispersion, and the third-order correction is $\frac{\ln n}{2n}$. Numerical results for the BPSK-modulated additive white Gaussian noise channel show that ORB-RCU remains close to the ML-RCU benchmark and that the resulting approximation accurately predicts finite-blocklength performance in the operating regime of practical interest.

cs.IT↗

arXiv2026

Scene-Conditioned PINN-GNN for Multipath RF Maps: Cross-Scene Generation and In-Scene Completion

Radio frequency (RF) maps provide a compact representation of multipath propagation characteristics and are fundamental to channel modeling, coverage analysis, and environment-aware wireless optimization. This paper proposes a unified RF map construction framework based on a physics-informed neural network (PINN) and a graph neural network (GNN), supporting both cross-scene generation and in-scene completion with 2D and 2.5D environmental representations. The PINN embeds electromagnetic propagation constraints to establish a physically consistent mapping from receiver locations to multipath parameters, including path gain, time of arrival, and angles, while the GNN enforces spatial consistency by modeling correlations among neighboring receivers. To comprehensively evaluate multipath reconstruction quality, we propose a peak-weighted dynamic time warping metric that jointly accounts for amplitude errors and peak delay misalignment in channel impulse responses. Extensive experiments demonstrate that the proposed method consistently outperforms image-based, diffusion-based, and interpolation baselines across both map-level and multipath-level metrics, achieving robust generalization and high-fidelity RF map construction under sparse observations.

eess.SP↗

arXiv2026

Change-Aware Self-Adaptive AI-Aided Kalman Filters With Neural Change Point Detection

Reliable state estimation in dynamical systems is often challenged by model mismatches, unknown noise statistics, and temporal variations. While AI-aided Kalman filters such as KalmanNet leverage deep learning to enhance classical estimation, they remain vulnerable to distribution shifts and lack mechanisms for autonomous adaptation. This work introduces Change-Aware Self-Adaptive KalmanNet (CASA-KalmanNet), an online adaptation framework that integrates a dedicated neural module, termed CPDNet, to monitor the interpretable internal features of KalmanNet and provide soft indicators of reliability degradation. These indicators dynamically regulate an online learning process, enabling data-efficient and timely adaptation to both abrupt and gradual changes in the system without requiring additional state labels from the changed regime. Numerical experiments on linear and nonlinear state-space models show that CASA-KalmanNet consistently outperforms existing learning-based filters under model mismatch, while approaching the accuracy of optimal classical methods with full domain knowledge.

eess.SY↗

arXiv2026

AeroVerse-SatAgent: UAV-Satellite Collaborative Spatial Reasoning Inspired by the Dual Visual Pathway Theory of Cognitive Neuroscience

With the rapid advancement of aerospace embodied intelligence, enabling Unmanned Aerial Vehicles (UAVs) to autonomously understand and reason about complex environments has become increasingly important. However, existing UAV-based spatial reasoning approaches face critical limitations: single-view perception renders them vulnerable to occlusions and perspective distortions, while most VLMs lack explicit geometric modeling, relying on semantic cues and yielding inconsistent reasoning under viewpoint and scale variations. To address these challenges, we propose SatAgent, a UAV-Satellite collaborative spatial reasoning model inspired by the dual-pathway mechanism of the human visual system. By jointly leveraging satellite and UAV perspectives, SatAgent enables robust, accurate reasoning in complex urban environments. We first introduce a Geometric-Aware 3D Reconstruction Encoder that elevates 2D UAV features into explicit 3D spatial representations. Next, we design a multi-view topology-semantic alignment module integrating cross-view features within a unified BEV coordinate system. We further introduce a multi-view consistency loss encouraging viewpoint-invariant representations. Finally, we construct SatAgent-SR130K, the first large-scale UAV-Satellite collaborative multi-view spatial reasoning dataset. Experiments show SatAgent outperforms state-of-the-art general-purpose foundation models and specialized spatial reasoning models by 25.91\% and 11.69\%, respectively, across diverse tasks, achieving particularly high accuracy in complex geometric relationship reasoning.

cs.CV↗

arXiv2026

A Gaussian Perspective for Distributional Discrepancy in Generative Diffusion Models

This paper introduces an analytical approach to quantifying and optimizing the distributional discrepancy in generative diffusion models. For a multivariate Gaussian source, we explicitly derive the closed-form evolution trajectory and the resulting Kullback-Leibler (KL) divergence between the distributions of the source data and the reversely sampled data. Asymptotic analysis via the Euler-Maclaurin expansion characterizes the convergence behavior of this KL divergence, extracting its dominant term as an explicit functional of the noise schedule. Minimizing this dominant term via the calculus of variations yields a noise schedule described by a tangent law, inherently determined by the source covariance spectrum. We further prove that the Gaussian source exhibits an extremal property for the KL divergence among general source distributions with a given covariance. We also utilize the analytical KL divergence as a principled metric to identify efficient time discretization strategies for pretrained diffusion models, and demonstrate via experiments over diverse datasets that the identified strategies consistently outperform established baselines, particularly under constrained function evaluation budgets.

cs.IT↗

arXiv2026

Performance Analysis and Optimal Design of ORB-Type GRAND Algorithms

Guessing Random Additive Noise Decoding (GRAND) performs decoding by sequentially guessing channel error patterns (EPs). Ordered Reliability Bits GRAND (ORBGRAND) is a notable instance suitable for efficient implementation, as it schedules EPs solely according to the ranking of soft channel outputs. In this paper, we generalize this principle to a broader class of GRAND algorithms whose testing order depends only on reliability ranking, referred to as ORB-type GRAND. We develop a unified analytical framework based on a key quantity termed the average guessing posterior (AGP), which captures the effectiveness of each EP and reduces decoding into an ordering problem over the EP space. For random code ensembles, we derive exact expressions for the block error rate (BLER), stopping-time distribution, and average number of tests under a fixed test budget. The analysis separates target-miss and target-preemption errors and shows that ordering EPs by non-increasing AGP is optimal over the EP set under consideration. For fixed linear block codes, we derive the BLER expression that isolates the code-dependent target-preemption term and characterize this term through higher-order weight relationships of codeword tuples, with a computable first-order upper bound as a useful special case. Guided by these insights, we formulate ReShuffled-ORBGRAND (RS-ORBGRAND) as an offline AGP-based reshuffling scheme. Numerical results for the Bose--Chaudhuri--Hocquenghem (BCH)$(127,113)$ code show that RS-ORBGRAND consistently improves existing ORB-type GRAND algorithms and lies within $0.1$~dB of a maximum-likelihood decoding lower-bound benchmark at a BLER of $10^{-6}$.

cs.IT↗

arXiv2026

A Survey of Learn-to-Compute Paradigms for Rate-Distortion-Type Problems

Rate-distortion (RD) theory and its related formulations play a central role in understanding efficient information representation, but computing these quantities remains challenging in high-dimensional settings. Classical iterative methods such as the Blahut-Arimoto algorithm become impractical in high-dimensional domains due to the curse of dimensionality and the intractability of mutual-information terms. Recent advances in neural modeling and differentiable optimization offer a promising alternative through a learn-to-compute paradigm, in which probability distributions and objective functionals are represented by flexible neural parameterizations. This survey presents an overview of neural approaches for evaluating the RD-type objectives. We present three representative families of methods: variational inference, neural mutual-information estimation, and dual-form optimization. By reviewing their theoretical principles, algorithmic techniques, and consistency properties, we elucidate how these methods collectively transform classical RD-type problems into scalable differentiable objectives suitable for deep learning, though challenges remain in large-scale applications. Together, these perspectives offer promising avenues for scaling information-theoretic computation to complex, high-dimensional machine learning systems.

cs.IT↗

arXiv2026

ORBGRAND Is Exactly Capacity-achieving via Rank Companding

Within the family of guessing-based decoding algorithms, ordered reliability bits GRAND (ORBGRAND) has attracted considerable attention due to its efficient use of soft information and suitability for hardware implementation. It has also been shown that ORBGRAND achieves a rate very close to the capacity of an additive white Gaussian noise channel under antipodal signaling. In this work, it is further established that, for general binary-input memoryless channels under symmetric input distribution, via suitably companding the ranks in ORBGRAND according to the inverse cumulative distribution function (CDF) of channel reliability, the resulting CDF-ORBGRAND algorithm exactly achieves the mutual information, i.e., the symmetric capacity. This result is then applied to bit-interleaved coded modulation (BICM) systems to handle high-order input constellations. Via considering the effects of mismatched decoding due to both BICM and ORBGRAND, it is shown that CDF-ORBGRAND is capable of achieving the BICM capacity, which was initially derived in the literature by treating BICM as a set of independent parallel channels.

cs.IT↗

arXiv2026

Peer-to-Peer Cloud Service Market for Data Centers Oriented to Computation-Electricity Coordination

Energy-intensive data centers (DCs) have emerged as substantial and flexible loads in modern power systems, underscoring the critical need for computation-electricity coordination. Harnessing the spatio-temporal flexibility of DC workloads is a promising approach to facilitate this coordination. However, existing studies overlook the collaborative potential of computational resource sharing among geo-distributed DCs, thereby failing to fully unlock this flexibility. In this paper, a bi-level computation-electricity coordination framework is proposed to explicitly capture the bidirectional interactions between DCs and power grid. Firstly, a peer-to-peer cloud service market (P2P-CSM) for geo-distributed DCs is proposed, which enables bilateral cloud service transactions to leverage regional heterogeneities (e.g., electricity prices, cooling efficiency). Secondly, locational marginal prices are embedded into the framework to reflect network congestion and nodal price disparities. Thirdly, a dual consensus alternating direction method of multipliers (ADMM)-based decentralized algorithm is developed as the P2P market clearing algorithm, and a bisection-assisted iterative algorithm is proposed to ensure rigorous convergence of the framework. Case studies conducted on modified IEEE 30-bus system validate that the P2P-CSM achieves a win-win computation-electricity coordination: it not only increases total DC operational profit by 22.8\%, but also effectively alleviates grid congestion and yields a 3.2\% reduction in total energy consumption.

eess.SY↗

arXiv2026

Sensing with Random Signals: The Role of Time Sharing

In monostatic, decision-aided, or known-waveform integrated sensing and communications (ISAC) formulations, the sensing receiver is often modeled as knowing the transmitted waveform. This assumption is not suitable for passive, bistatic, or distributed settings where the sensing receiver knows the signaling rule but not the transmitted symbols. We study such a symbol-unaware ISAC model, where sensing is measured by the unconditioned mutual information $I(S;V)$ rather than the symbol-aware quantity $I(S;V|X)$. For discrete-input memoryless channels, we characterize the capacity-sensing region through an auxiliary time-sharing variable, showing that the optimal upper boundary is the upper concave envelope of the single-mode frontier. Thus, explicit time sharing is unnecessary when the single-mode frontier is already concave, but strictly beneficial when its upper concave envelope strictly dominates the frontier. For Rayleigh-fading BPSK, we further show that the curvature of the single-mode boundary is determined by the stochastic ordering of the communication- and sensing-side effective SNR distributions. Communication-side dominance yields a concave single-mode frontier and no time-sharing gain, sensing-side dominance yields a convex single-mode frontier and a strict time-sharing gain, and equality yields a linear boundary. The result extends to SIMO-BPSK through the ordering of post-combining SNR distributions. These findings explain when symbol-unaware ISAC optimally moves from data-symbol transmission to pilot-like sensing modes.

cs.IT↗

arXiv2026

Vectorized Generalized Nearest Neighbor Decoding for In-block Memory Channel

This work extends the generalized nearest neighbor decoding (GNND), originally developed as a receiver architecture for memoryless channels, to a vectorized GNND (Vec-GNND) suitable for in-block memory (IBM) channels. Leveraging the generalized mutual information (GMI) as an operational lower bound on the mismatch capacity, an analytical characterization of the optimal Vec-GNND is obtained for general IBM channels with Gaussian codebooks. The formalism further provides closed-form optimality conditions and achievable GMIs for restricted variants of the receiver architecture. Furthermore, we formulate a GMI-based joint design viewpoint for Gaussian codebook covariance and decoding metrics. Since the metric optimization admits a closed-form solution for each fixed covariance, the joint design is reduced to an input-covariance optimization problem; for the diagonal covariance family, we derive first-order self-consistent optimality conditions. Numerical evaluations on block noncoherent additive white Gaussian noise channels and phase noise channels demonstrate consistent performance gains over conventional scaling-based baselines, highlighting the substantial advantages and potential relevance of the proposed Vec-GNND in realistic communication scenarios.

cs.IT↗

arXiv2026

Relative Score Policy Optimization for Diffusion Language Models

Diffusion large language models (dLLMs) offer a promising route to parallel and efficient text generation, but improving their reasoning ability requires effective post-training. Reinforcement learning with verifiable rewards (RLVR) is a natural choice for this purpose, yet its application to dLLMs is hindered by the absence of tractable sequence-level log-ratios, which are central to standard policy optimization. The lack of tractable sequence-level log-ratios forces existing methods to rely on high-variance ELBO-based approximations, where high verifier rewards can amplify inaccurate score estimates and destabilize RL training. To overcome this issue, we propose \textbf{R}elative \textbf{S}core \textbf{P}olicy \textbf{O}ptimization (RSPO), a simple RLVR method that uses verifiable rewards to calibrate noisy likelihood estimates in dLLMs. The core of our algorithm relies on a key observation: a reward advantage can be interpreted not only as an update direction, but also as a target for the relative log-ratio between the current and reference policies. Accordingly, RSPO calibrates this noisy relative log-ratio estimate by comparing its reward advantage with the reward-implied target relative log-ratio, updating the policy according to the gap between the current estimate and the target rather than the raw advantage alone. Experiments on mathematical reasoning and planning benchmarks show that RSPO yields especially strong gains on planning tasks and competitive mathematical-reasoning performance.

cs.CL↗

arXiv2026

A Parallelization Strategy for GRAND with Optimality Guarantee by Exploiting Error Pattern Tree Representation

Parallelism has become a central concern in modern decoding frameworks aiming to meet stringent throughput and latency requirements. Guessing Random Additive Noise Decoding (GRAND) is a recently proposed decoding paradigm that tests candidate Error Patterns (EPs) until a valid codeword is found. Among its variants, Soft GRAND (SGRAND) achieves maximum-likelihood (ML) decoding but relies on real-time generation and likelihood ordering of EPs, making parallel execution nontrivial under the ML optimality constraint. In this work, we introduce a unified binary tree representation of EPs, termed the EP tree, which formalizes the hierarchical structure underlying SGRAND and Ordered Reliability Bits (ORB) GRAND algorithms, enabling structured organization of EPs and algorithmic-level parallel exploration. Building upon this unified framework, we propose a parallel design of SGRAND that preserves ML optimality while significantly reducing decoding complexity through pruning strategies and tree-based computation. Furthermore, we develop an enhanced ORBGRAND algorithm based on the same EP tree representation, improving decoding performance toward ML while retaining parallel efficiency. Numerical experiments show that the proposed parallel SGRAND achieves a $3.96\times$ reduction in decoding latency compared with its serial counterpart, while the enhanced ORBGRAND achieves a $4.21\times$ speedup, demonstrating the effectiveness of the unified tree-based framework and its strong potential for future algorithmic and hardware optimizations.

cs.IT↗

arXiv2026

A Remark on Downlink Massive Random Access

In downlink massive random access (DMRA), a base station transmits messages to a typically small subset of active users, selected randomly from a massive number of total users. Explicitly encoding the identities of active users would incur a significant overhead scaling logarithmically with the number of total users. Recently, via a random coding argument, Song, Attiah and Yu have shown that the overhead can be reduced to within some upper bound irrespective of the number of total users. In this remark, recognizing that the code design for DMRA is an instance of covering arrays in combinatorics, we show that there exists deterministic construction of variable-length codes that incur an overhead no greater than $1 + log_2 e$ bits.

cs.IT↗

arXiv2026

PINN and GNN-based RF Map Construction for Wireless Communication Systems

Radio frequency (RF) map is a promising technique for capturing the characteristics of multipath signal propagation, offering critical support for channel modeling, coverage analysis, and beamforming in wireless communication networks. This paper proposes a novel RF map construction method based on a combination of physics-informed neural network (PINN) and graph neural network (GNN). The PINN incorporates physical constraints derived from electromagnetic propagation laws to guide the learning process, while the GNN models spatial correlations among receiver locations. By parameterizing multipath signals into received power, delay, and angle of arrival (AoA), and integrating both physical priors and spatial dependencies, the proposed method achieves accurate prediction of multipath parameters. Experimental results demonstrate that the method enables high-precision RF map construction under sparse sampling conditions and delivers robust performance in both indoor and complex outdoor environments, outperforming baseline methods in terms of generalization and accuracy.

eess.SP↗

arXiv2026

OpenPathNet: An Open-Source RF Multipath Data Generator for AI-Driven Wireless Systems

The convergence of artificial intelligence (AI) and sixth-generation (6G) wireless technologies is driving an urgent need for large-scale, high-fidelity, and reproducible radio frequency (RF) datasets. Existing resources, such as CKMImageNet, primarily provide preprocessed and image-based channel representations, which conceal the fine-grained physical characteristics of signal propagation that are essential for effective AI modeling. To bridge this gap, we present OpenPathNet, an open-source RF multipath data generator accompanied by a publicly released dataset for AI-driven wireless research. Distinct from prior datasets, OpenPathNet offers disaggregated and physically consistent multipath parameters, including per-path gain, time of arrival (ToA), and spatial angles, derived from high-precision ray tracing simulations constructed on real-world environment maps. By adopting a modular, parameterized pipeline, OpenPathNet enables reproducible generation of multipath data and can be readily extended to new environments and configurations, improving scalability and transparency. The released generator and accompanying dataset provide an extensible testbed that holds promise for advancing studies on channel modeling, beam prediction, environment-aware communication, and integrated sensing in AI-enabled 6G systems. The source code and dataset are publicly available at https://github.com/liu-lz/OpenPathNet.

eess.SP↗

arXiv2026

Detecting Transportation Mode Using Dense Smartphone GPS Trajectories and Transformer Models

Transportation mode detection is an important topic within GeoAI and transportation research. In this study, we introduce SpeedTransformer, a novel Transformer-based model that relies solely on speed inputs to infer transportation modes from dense smartphone GPS trajectories. In benchmark experiments, SpeedTransformer outperformed traditional deep learning models, such as the Long Short-Term Memory (LSTM) network. Moreover, the model demonstrated strong flexibility in transfer learning, achieving high accuracy across geographical regions after fine-tuning with small datasets. Finally, we deployed the model in a real-world experiment, where it consistently outperformed baseline models under complex built environments and high data uncertainty. These findings suggest that Transformer architectures, when combined with dense GPS trajectories, hold substantial potential for advancing transportation mode detection and broader mobility-related research.

cs.LG↗