arXiv subjects

Hanxi Fang

Publications and source records attributed to Hanxi Fang.

6 recordsLinked to original sources

arXiv2026

LazyAttention: Efficient Retrieval-Augmented Generation with Deferred Positional Encoding

Key-value (KV) caching accelerates inference of large language models (LLMs) by reusing past computations for generated tokens. Its importance becomes even greater in long-context applications such as retrieval-augmented generation (RAG) and in-context learning (ICL). However, conventional KV caching embeds positional information directly into the cache, limiting its reusability. Existing solutions either restrict reuse to prefixes or require expensive memory materialization for positional re-encoding. We introduce LazyAttention, a novel attention mechanism that kernelizes deferred positional encoding to enable zero-copy, position-agnostic KV reuse. By adjusting positional encoding within attention kernels on-the-fly, LazyAttention resolves the materialization bottleneck, allowing a single physical KV copy to serve multiple logical requests at arbitrary positions. Leveraging attention kernels tailored for prefilling and decoding, our system achieves significant efficiency improvements: under skewed document distributions, it reduces time-to-first-token (TTFT) by 1.37$\times$ and increases inference throughput by 1.40$\times$ compared to the state-of-the-art Block-Attention, while maintaining comparable output quality.

cs.CL↗

arXiv2025

Enhancing Computational Notebooks with Code+Data Space Versioning

There is a gap between how people explore data and how Jupyter-like computational notebooks are designed. People explore data nonlinearly, using execution undos, branching, and/or complete reverts, whereas notebooks are designed for sequential exploration. Recent works like ForkIt are still insufficient to support these multiple modes of nonlinear exploration in a unified way. In this work, we address the challenge by introducing two-dimensional code+data space versioning for computational notebooks and verifying its effectiveness using our prototype system, Kishuboard, which integrates with Jupyter. By adjusting code and data knobs, users of Kishuboard can intuitively manage the state of computational notebooks in a flexible way, thereby achieving both execution rollbacks and checkouts across complex multi-branch exploration history. Moreover, this two-dimensional versioning mechanism can easily be presented along with a friendly one-dimensional history. Human subject studies indicate that Kishuboard significantly enhances user productivity in various data science tasks.

cs.HC↗

arXiv2025

Large-scale Evaluation of Notebook Checkpointing with AI Agents

Saving, or checkpointing, intermediate results during interactive data exploration can potentially boost user productivity. However, existing studies on this topic are limited, as they primarily rely on small-scale experiments with human participants - a fundamental constraint of human subject studies. To address this limitation, we employ AI agents to simulate a large number of complex data exploration scenarios, including revisiting past states and branching into new exploration paths. This strategy enables us to accurately assess the impact of checkpointing while closely mimicking the behavior of real-world data practitioners. Our evaluation results, involving more than 1,000 exploration paths and 2,848 executed code blocks, show that a checkpointing framework for computational notebooks can indeed enhance productivity by minimizing unnecessary code re-executions and redundant variables or code.

cs.HC↗

arXiv2023

Forecasting Inter-Destination Tourism Flow via a Hybrid Deep Learning Model

Tourists often go to multiple tourism destinations in one trip. The volume of tourism flow between tourism destinations, also referred to as ITF (Inter-Destination Tourism Flow) in this paper, is commonly used for tourism management on tasks like the classification of destinations' roles and visitation pattern mining. However, the ITF is hard to get due to the limitation of data collection techniques and privacy issues. It is difficult to understand how the volume of ITF is influenced by features of the multi-attraction system. To address these challenges, we utilized multi-source datasets and proposed a graph-based hybrid deep learning model to predict the ITF. The model makes use of both the explicit features of individual tourism attractions and the implicit features of the interactions between multiple attractions. Experiments on ITF data extracted from crowdsourced tourists' travel notes about the city of Beijing verified the usefulness of the proposed model. Besides, we analyze how different features of tourism attractions influence the volume of ITF with explainable AI techniques. Results show that popularity, quality and distance are the main three influential factors. Other features like coordinates will also exert an influence in different ways. The predicted ITF data can be further used for various downstream tasks in tourism management. The research also deepens the understanding of tourists' visiting choice in a tourism system consisting of multiple attractions.

cs.CY↗

arXiv2023

Spatio-Temporal Trajectory Similarity Measures: A Comprehensive Survey and Quantitative Study

Spatio-temporal trajectory analytics is at the core of smart mobility solutions, which offers unprecedented information for diversified applications such as urban planning, infrastructure development, and vehicular networks. Trajectory similarity measure, which aims to evaluate the distance between two trajectories, is a fundamental functionality of trajectory analytics. In this paper, we propose a comprehensive survey that investigates all the most common and representative spatio-temporal trajectory measures. First, we provide an overview of spatio-temporal trajectory measures in terms of three hierarchical perspectives: Non-learning vs. Learning, Free Space vs. Road Network, and Standalone vs. Distributed. Next, we present an evaluation benchmark by designing five real-world transformation scenarios. Based on this benchmark, extensive experiments are conducted to study the effectiveness, robustness,nefficiency, and scalability of each measure, which offers guidelines for trajectory measure selection among multiple techniques and applications such as trajectory data mining, deep learning, and distributed processing.

cs.DS↗

arXiv2022

Estimator: An Effective and Scalable Framework for Transportation Mode Classification over Trajectories

Transportation mode classification, the process of predicting the class labels of moving objects transportation modes, has been widely applied to a variety of real world applications, such as traffic management, urban computing, and behavior study. However, existing studies of transportation mode classification typically extract the explicit features of trajectory data but fail to capture the implicit features that affect the classification performance. In addition, most of the existing studies also prefer to apply RNN-based models to embed trajectories, which is only suitable for classifying small-scale data. To tackle the above challenges, we propose an effective and scalable framework for transportation mode classification over GPS trajectories, abbreviated Estimator. Estimator is established on a developed CNN-TCN architecture, which is capable of leveraging the spatial and temporal hidden features of trajectories to achieve high effectiveness and efficiency. Estimator partitions the entire traffic space into disjointed spatial regions according to traffic conditions, which enhances the scalability significantly and thus enables parallel transportation classification. Extensive experiments using eight public real-life datasets offer evidence that Estimator i) achieves superior model effectiveness (i.e., 99% Accuracy and 0.98 F1-score), which outperforms state-of-the-arts substantially; ii) exhibits prominent model efficiency, and obtains 7-40x speedups up over state-of-the-arts learning-based methods; and iii) shows high model scalability and robustness that enables large-scale classification analytics.

cs.LG↗