Large language model (LLM) agents have emerged as a promising solution for enhancing recommendation systems via user simulation. However, existing studies predominantly resort to prompt-based simulation using frozen LLMs, which frequently results in suboptimal item modeling and user preference learning, thereby ultimately constraining recommendation performance. To address these challenges, we introduce VRAgent-R1, a novel agent-based paradigm that incorporates human-like intelligence in user simulation. Specifically, VRAgent-R1 comprises two distinct agents: the Item Perception (IP) Agent and the User Simulation (US) Agent, designed for interactive user-item modeling. Firstly, the IP Agent emulates human-like progressive thinking based on MLLMs, effectively capturing hidden recommendation semantics in videos. With a more comprehensive multimodal content understanding provided by the IP Agent, the video recommendation system is equipped to provide higher-quality candidate items. Subsequently, the US Agent refines the recommended video sets based on in-depth chain-of-thought (CoT) reasoning and achieves better alignment with real user preferences through reinforcement learning. Experimental results on a large-scale video recommendation benchmark MicroLens-100k have demonstrated the effectiveness of our proposed VRAgent-R1 method, e.g., the IP Agent achieves a 6.0% improvement in NDCG@10, while the US Agent shows approximately 45.0% higher accuracy in user decision simulation compared to state-of-the-art baselines.

With the prevalence of Multimodal Large Language Models(MLLMs), autonomous driving has encountered new opportunities and challenges. In particular, multi-modal video understanding is critical to interactively analyze what will happen in the procedure of autonomous driving. However, videos in such a dynamical scene that often contains complex spatial-temporal movements, which restricts the generalization capacity of the existing MLLMs in this field. To bridge the gap, we propose a novel Hierarchical Mamba Adaptation (H-MBA) framework to fit the complicated motion changes in autonomous driving videos. Specifically, our H-MBA consists of two distinct modules, including Context Mamba (C-Mamba) and Query Mamba (Q-Mamba). First, C-Mamba contains various types of structure state space models, which can effectively capture multi-granularity video context for different temporal resolution. Second, Q-Mamba flexibly transforms the current frame as the learnable query, and attentively select multi-granularity video context into query. Consequently, it can adaptively integrate all the video contexts of multi-scale temporal resolutions to enhance video understanding. Via a plug-and-play paradigm in MLLMs, our H-MBA shows the remarkable performance on multi-modal video tasks in autonomous driving, e.g., for risk object detection, it outperforms the previous SOTA method with 5.5% mIoU improvement.

Multi-city human mobility prediction is critical for GIScience and urban computing. However, pronounced spatiotemporal heterogeneity and training instability on large-scale, multi-city datasets restrict model generalization. To address these challenges, we propose the Human Mobility Predictor (HuMob Predictor), a novel framework that enhances multi-city mobility prediction. Our model introduces a multi-level encoding module that captures cross-city heterogeneity and within-city spatial relationships. City encodings capture macro-level characteristics unique to each city, while absolute spatial encodings preserve geographic proximity across grid cells. By combining these encodings, HuMob Predictor learns universal mobility representations while retaining city-specific features. To stabilize training across cities, we adopt an incremental training strategy that gradually increases prediction difficulty, significantly improving convergence and cross-city generalization. Experiments on the GISCUP 2025 multi-city datasets demonstrate that HuMob Predictor achieves superior performance in individual mobility prediction.

Understanding and simulating individual human mobility within intra-urban areas are essential for urban planning, transportation, public health, and other related fields. Currently, Random Walk-based and Exploration and Preferential Return (EPR)-based models are the primary mechanistic approaches used to model individual human mobility. However, these models often rely on assumptions and parameters derived from studies conducted at larger spatial scales, such as the national level. This results in inaccuracies when applied to urban contexts because individual mobility at the intra-urban scale exhibits unique spatiotemporal characteristics. Through a systematic analysis of three consecutive trajectory datasets and three check-in trajectory datasets collected from six cities worldwide, we identified two key insights into intra-urban mobility: (1) individuals’ waiting time distributions follow a log-normal function with universal parameters, rather than the gamma, power-law, or exponential functions commonly used in current studies, and (2) individuals’ spatial choice behavior is better captured by the Universal Opportunity (UO) model with universal parameters, rather than the frequently used gravity models. Based on these findings, we proposed Urban-EPR, a universal model specifically designed to simulate intra-urban individual mobility. Additionally, we compared seven mainstream mechanistic models, demonstrating the superior performance of Urban-EPR and providing a benchmark for intra-urban mobility research.

Knowledge Distillation (KD) transfers knowledge from large models to small models and has recently achieved remarkable success. However, the reliability of existing KD methods in real-world applications, especially under distribution shift, remains underexplored. Distribution shift refers to the data distribution drifts between the training and testing phases, and this can adversely affect the efficacy of KD. In this paper, we propose a unified and systematic framework ShiftKD to benchmark KD against two general distributional shifts: diversity and correlation shift. The evaluation benchmark covers more than 30 methods from algorithmic, data-driven, and optimization perspectives for five benchmark datasets. Our development of ShiftKD conducts extensive experiments and reveals strengths and limitations of current SOTA KD methods. More importantly, we thoroughly analyze key factors in student model training process, including data augmentation, pruning methods, optimizers, and evaluation metrics. We believe ShiftKD could serve as an effective benchmark for assessing KD in real-world scenarios, thus driving the development of more robust KD methods in response to evolving demands. The code will be made available upon publication.

Understanding human mobility patterns is essential for various applications, from urban planning to public safety. The individual trajectory such as mobile phone location data, while rich in spatio-temporal information, often lacks semantic detail, limiting its utility for in-depth mobility analysis. Existing methods can infer basic routine activity sequences from this data, lacking depth in understanding complex human behaviors and users’ characteristics. Additionally, they struggle with the dependency on hard-to-obtain auxiliary datasets like travel surveys. To address these limitations, this paper defines trajectory semantic inference through three key dimensions: user occupation category, activity sequence, and trajectory description, and proposes the Trajectory Semantic Inference with Large Language Models (TSI-LLM) framework to leverage LLMs infer trajectory semantics comprehensively and deeply. We adopt spatio-temporal attributes enhanced data formatting (STFormat) and design a context-inclusive prompt, enabling LLMs to more effectively interpret and infer the semantics of trajectory data. Experimental validation on real-world trajectory datasets demonstrates the efficacy of TSI-LLM in deciphering complex human mobility patterns. This study explores the potential of LLMs in enhancing the semantic analysis of trajectory data, paving the way for more sophisticated and accessible human mobility research.