arXiv subjects

Chengbo Zhang

Publications and source records attributed to Chengbo Zhang.

7 recordsLinked to original sources

arXiv2026

Noise-driven pseudovorticity multipoles in self-focusing beams with quintic saturation

We investigate pseudovorticity generation in Gaussian beams undergoing self-focusing under amplitude and phase noise, using the cubic-quintic nonlinear Schr\"odinger equation. Pseudovorticity, defined as the curl of the optical momentum flux, characterizes local rotational flow in the absence of phase singularities. Our numerical simulations show that thermal amplitude and phase noise induce a multipolar pseudovorticity pattern. Unlike the pure cubic case, where noise asymmetries are radiated away during collapse, the quintic saturation arrests collapse and traps the noise in the resulting soliton. Hence, pseudovorticity multipoles persist, oscillating at the focusing-refocusing period. These results suggest a potential pathway for controlling local optical torque through noise engineering.

physics.optics↗

arXiv2026

A Knowledge-Grounded Behavioral Reasoning Framework for Training-Free Urban Healthcare OD Prediction

Understanding urban healthcare mobility is essential for healthcare resource allocation, hospital capacity planning, and resilient urban governance. Existing healthcare origin-destination (OD) prediction methods primarily rely on supervised deep learning, requiring expensive model training while providing limited interpretability into the behavioral mechanisms of hospital choice. This paper presents a training-free urban healthcare OD prediction framework based on knowledge-grounded behavioral reasoning, replacing gradient-based neural network learning with collaborative LLM reasoning over structured urban knowledge. The framework organizes heterogeneous urban information, including healthcare resources, population characteristics, transportation accessibility, weather conditions, and historical mobility, into a unified knowledge graph, enabling evidence retrieval, urban context understanding, behavioral reasoning, preference ranking, and Bayesian verification through a multi-agent reasoning pipeline. Experiments on a real-world Shenzhen dataset containing 137 hospitals and 6,341 prediction tasks demonstrate that the proposed framework achieves CPC = 0.5449, Top-1 = 66.19%, Top-5 = 68.03%, and Top-10 = 74.03%, consistently outperforming conventional deep learning baselines on all Top-K metrics (e.g., Top-5 improves by 22.1% over MLP). Compared with Direct LLM using the same backbone without structured reasoning, the proposed framework improves CPC by 87.5% and Top-1 accuracy by 141%, demonstrating that knowledge-grounded behavioral reasoning, rather than raw LLM capability, is the key to accurate, interpretable, and training-free healthcare OD prediction. These findings highlight the potential of urban intelligence to move healthcare mobility modeling beyond data fitting toward knowledge-driven behavioral reasoning.

physics.soc-ph↗

arXiv2026

Large Language Models as Delivery Rider: Generating Instant Food Delivery Riders' Routing Decision with LLM Agent Framework

The utilization of Large Language Models (LLMs) to power human-like agents has shown remarkable potential in simulating individual mobility pattern. However, a significant gap remains in modeling cohorts of agents in dynamic and interactive systems where they must take strategic routing decisions to response mobility-specific task. To bridge this gap, we introduce LLM-DR, a novel agent framework designed to simulate the heterogeneous decision-making of riders in the on-demand instant delivery task scenario. Our framework is founded on two principles: 1) Empirically-grounded personas, where we use unsupervised clustering on a large-scale, real-world trajectory dataset to identify four distinct rider work strategies; and 2) Reasoning-based routing process, where each persona is instantiated as an LLM agent that employs a structured Chain-of-Thought (CoT) process to make human-like routing choices. This framework enables the construction of high-fidelity simulations to investigate how the strategic composition of a rider workforce influences system-level outcomes regarding their mobility pattern. We validate our framework on an real-world instant deliver order datasets, demonstrating its capacity to model complex rider behavior in an interactive market scenario. This work provides pioneering findings in agentic mobility system empowered by LLM.

physics.soc-ph↗

arXiv2025

Deciphering Delivery Mobility: A City-Scale, Path-Reconstructed Trajectory Dataset of Instant Delivery Riders

The rapid expansion of the on-demand economy has profoundly reshaped urban mobility and logistics, yet open data linking multi-stop delivery task sequences with route geometry remain scarce. Here, we present a city-scale, path-reconstructed route dataset for instant-delivery tasks in Beijing, built from desensitized platform records from Ele.me. The dataset contains 79,648 reconstructed delivery-wave records associated with 986 anonymized courier identifiers and covering 267,529 orders during February 2020. For each wave, Assign, Pickup, and Delivery action points are ordered by source-record timestamps, and Amap cycling routes are queried between consecutive action points to create continuous route geometries. The resulting dataset records navigation-based riding paths and the associated action sequence for each delivery wave, revealing insightful spatiotemporal regularities. Source-record validation shows strong agreement for distance and moderate agreement for duration. This urban data resource enables researchers in urban analytics and transportation management to investigate delivery activities, model urban logistics systems, and develop sustainable policies.

physics.soc-ph↗

arXiv2025

STORM-BORN: A Challenging Mathematical Derivations Dataset Curated via a Human-in-the-Loop Multi-Agent Framework

High-quality math datasets are crucial for advancing the reasoning abilities of large language models (LLMs). However, existing datasets often suffer from three key issues: outdated and insufficient challenging content, neglecting human-like reasoning, and limited reliability due to single-LLM generation. To address these, we introduce STORM-BORN, an ultra-challenging dataset of mathematical derivations sourced from cutting-edge academic papers, which includes dense human-like approximations and heuristic cues. To ensure the reliability and quality, we propose a novel human-in-the-loop, multi-agent data generation framework, integrating reasoning-dense filters, multi-agent collaboration, and human mathematicians' evaluations. We curated a set of 2,000 synthetic samples and deliberately selected the 100 most difficult problems. Even most advanced models like GPT-o1 solved fewer than 5% of them. Fine-tuning on STORM-BORN boosts accuracy by 7.84% (LLaMA3-8B) and 9.12% (Qwen2.5-7B). As AI approaches mathematician-level reasoning, STORM-BORN provides both a high-difficulty benchmark and a human-like reasoning training resource. Our code and dataset are publicly available at https://github.com/lwhere/STORM-BORN.

cs.CL↗

arXiv2024

Enhancing Multi-level Urban Instant Delivery Management via Infomap-based Hierarchical Community Detection

Efficient management of on-demand delivery systems is essential for modern urban logistics, especially in densely populated cities with complex spatial layouts. This study introduces a novel, computer-supported cooperative framework that utilizes Infomap-based hierarchical community detection to analyze spatial multilevel clustering patterns. The experiment was conducted to large scale on-demand delivery datasets from Shenzhen and Beijing, revealing integrated spatial clusters that align with cohesive urban layout. Through hierarchical detection, finer and fragmented clusters are identified, reflecting its diverse urban structure and delivery demands. The findings demonstrate the effectiveness of hierarchical community detection in uncovering spatial dependencies and optimizing resource allocation and delivery strategies. This framework provides practical insights for urban logistics, enabling tailored approaches for business hub placement, route allocation, and adaptive resource management.

physics.soc-ph↗

arXiv2024

Uncover the Dynamic Community Structure of Instant Delivery Network

The rise of instant delivery services has reshaped urban spatial structures through the interaction between suppliers and consumers. However, limited research has explored the spatiotemporal dynamics of delivery network structures. This study constructs a time-dependent, multi-layer instant delivery network in the case city of Beijing using a large-scale dataset from Eleme, organized into 500m grid units. A dynamic community detection method identifies evolving community structures over time. The results reveal 309 dynamic communities, with an average size of 13.78 square kilometers. Communities form in the morning, expand, stabilize, then contract, and disappear by night. Key factors influencing stability include building area and residential population, while online retail and service facilities contribute to instability. These findings offer insights into the spatial structure of instant delivery networks and the factors driving their dynamics, with practical implications for optimizing platform strategies, resource allocation, and urban transportation planning.

physics.soc-ph↗