arXiv subjects

Wenwei Li

Publications and source records attributed to Wenwei Li.

8 recordsLinked to original sources

arXiv2026

Pushing Bistatic Wireless Sensing toward High Accuracy at the Sub-Wavelength Scale

Contactless sensing using wireless communication signals has garnered significant attention due to its non-intrusive nature and ubiquitous infrastructure. Despite the promise, the inherent bistatic deployment of wireless communication introduces clock asynchronism, which leads to unknown phase offsets in channel response and hinders fine-grained sensing. State-of-the-art systems widely adopt the cross-antenna channel ratio to cancel these detrimental phase offsets. However, the channel ratio preserves sensing feature accuracy only at integer-wavelength target displacements, losing sub-wavelength fidelity. To overcome this limitation, we derive the first quantitative mapping between the distorted ratio feature and the ideal channel feature. Building on this foundation, we develop a robust framework that leverages channel response amplitude to recover the ideal channel feature from the distorted ratio. Real-world experiments across Wi-Fi and LoRa demonstrate that our method can effectively reconstruct sub-wavelength displacement details, achieving nearly an order-of-magnitude improvement in accuracy.

cs.IT↗

arXiv2026

Search-P1: Path-Centric Reward Shaping for Stable and Efficient Agentic RAG Training

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, yet traditional single-round retrieval struggles with complex multi-step reasoning. Agentic RAG addresses this by enabling LLMs to dynamically decide when and what to retrieve, but current RL-based training methods suffer from sparse outcome rewards that discard intermediate signals and low sample efficiency where failed samples contribute nothing. We propose Search-P1, a framework that introduces path-centric reward shaping for agentic RAG training, comprising two key components: (1) Path-Centric Reward, which evaluates the structural quality of reasoning trajectories through order-agnostic step coverage and soft scoring that extracts learning signals even from failed samples, and (2) Dual-Track Path Scoring with offline-generated reference planners that assesses paths from both self-consistency and reference-alignment perspectives. Experiments on multiple QA benchmarks demonstrate that Search-P1 achieves significant improvements over Search-R1 and other strong baselines, with an average accuracy gain of 7.7 points.

cs.CL↗

arXiv2026

Towards Faithful Industrial RAG: A Reinforced Co-adaptation Framework for Advertising QA

Industrial advertising question answering (QA) is a high-stakes task in which hallucinated content, particularly fabricated URLs, can lead to financial loss, compliance violations, and legal risk. Although Retrieval-Augmented Generation (RAG) is widely adopted, deploying it in production remains challenging because industrial knowledge is inherently relational, frequently updated, and insufficiently aligned with generation objectives. We propose a reinforced co-adaptation framework that jointly optimizes retrieval and generation through two components: (1) Graph-aware Retrieval (GraphRAG), which models entity-relation structure over a high-citation knowledge subgraph for multi-hop, domain-specific evidence selection; and (2) evidence-constrained reinforcement learning via Group Relative Policy Optimization (GRPO) with multi-dimensional rewards covering faithfulness, style compliance, safety, and URL validity. Experiments on an internal advertising QA dataset show consistent gains across expert-judged dimensions including accuracy, completeness, and safety, while reducing the hallucination rate by 72\%. A two-week online A/B test demonstrates a 28.6\% increase in like rate, a 46.2\% decrease in dislike rate, and a 92.7\% reduction in URL hallucination. The system has been running in production for over half a year and has served millions of QA interactions.

cs.CL↗

arXiv2026

AD-Bench: A Real-World, Trajectory-Aware Advertising Analytics Benchmark for LLM Agents

While Large Language Model (LLM) agents have made remarkable progress on complex reasoning, evaluating them in real-world environments remains an open problem. Existing benchmarks are largely confined to idealized simulations and fail to capture specialized domains such as advertising and marketing analytics, where tasks require multi-round interaction with professional tools and where ground-truth answers quickly become obsolete as data and platform rules evolve. To address this, we propose AD-Bench, a benchmark built from real user marketing-analysis requests on a production advertising platform. AD-Bench introduces two key designs: (i) a dynamic ground-truth pipeline that replays expert tool-call trajectories to regenerate answers consistent with the current environment, mitigating answer obsolescence; and (ii) a trajectory-aware evaluation that jointly measures end-to-end answer correctness (Pass@k) and trajectory coverage. Requests are stratified into three difficulty levels (L1-L3) to probe multi-round, multi-tool collaboration. Experiments show that the best model, Claude-Opus-4.7, attains Pass@1 = 76.9% and Pass@3 = 80.4% with 82.7% trajectory coverage overall, yet drops sharply on L3 to Pass@1 = 61.4% and Pass@3 = 65.1%, revealing that even state-of-the-art agents have substantial gaps in complex advertising analytics.

cs.CL↗

arXiv2025

Semantic-Enhanced Feature Matching with Learnable Geometric Verification for Cross-Modal Neuron Registration

Accurately registering in-vivo two-photon and ex-vivo fluorescence micro-optical sectioning tomography images of individual neurons is critical for structure-function analysis in neuroscience. This task is profoundly challenging due to a significant cross-modality appearance gap, the scarcity of annotated data and severe tissue deformations. We propose a novel deep learning framework to address these issues. Our method introduces a semantic-enhanced hybrid feature descriptor, which fuses the geometric precision of local features with the contextual robustness of a vision foundation model DINOV3 to bridge the modality gap. To handle complex deformations, we replace traditional RANSAC with a learnable Geometric Consistency Confidence Module, a novel classifier trained to identify and reject physically implausible correspondences. A data-efficient two-stage training strategy, involving pre-training on synthetically deformed data and fine-tuning on limited real data, overcomes the data scarcity problem. Our framework provides a robust and accurate solution for high-precision registration in challenging biomedical imaging scenarios, enabling large-scale correlative studies.

eess.IV↗

arXiv2025

A Few-Shot Metric Learning Method with Dual-Channel Attention for Cross-Modal Same-Neuron Identification

In neuroscience research, achieving single-neuron matching across different imaging modalities is critical for understanding the relationship between neuronal structure and function. However, modality gaps and limited annotations present significant challenges. We propose a few-shot metric learning method with a dual-channel attention mechanism and a pretrained vision transformer to enable robust cross-modal neuron identification. The local and global channels extract soma morphology and fiber context, respectively, and a gating mechanism fuses their outputs. To enhance the model's fine-grained discrimination capability, we introduce a hard sample mining strategy based on the MultiSimilarityMiner algorithm, along with the Circle Loss function. Experiments on two-photon and fMOST datasets demonstrate superior Top-K accuracy and recall compared to existing methods. Ablation studies and t-SNE visualizations validate the effectiveness of each module. The method also achieves a favorable trade-off between accuracy and training efficiency under different fine-tuning strategies. These results suggest that the proposed approach offers a promising technical solution for accurate single-cell level matching and multimodal neuroimaging integration.

cs.CV↗

arXiv2018

Approximation of the Partition Number After Hardy and Ramanujan: An Application of Data Fitting Method in Combinatorics

Sometimes we need the approximate value of the partition number in a simple and efficient way. There are already several formulae to calculate the partition number p(n). But they are either inconvenient for most people (not majored in math) who do not want do write programs, or unsatisfying in accuracy. By bringing in two parameters in the Hardy-Ramanujan's Asymptotic formula and fitting the data of the two parameters by least square method, iteration method and some other special designed methods, several revised elementary estimation formulae with high accuracy for p(n) are obtained. With these estimation formulae, the approximate value of p(n) can be calculated by a pocket calculator without programming function. The main difficulty is that the usual methods to fit the data of the two parameters by an elementary function is defective here. These method could be used in finding the fitting functions of some other complex data.

math.NT↗

arXiv2018

A Note on the Classification of Permutation Matrix

This paper is concentrated on the classification of permutation matrix with the permutation similarity relation, mainly about the canonical form of a permutational similar equivalence class, the cycle matrix decomposition of a permutation matrix and the cycle factorization of a permutation matrix or monomial matrix.

math.GM↗