Zeshui Xu
YOU?
Author Swipe
View article: RLinf-VLA: A Unified and Efficient Framework for VLA+RL Training
RLinf-VLA: A Unified and Efficient Framework for VLA+RL Training Open
Recent progress in vision and language foundation models has significantly advanced multimodal understanding, reasoning, and generation, inspiring a surge of interest in extending such capabilities to embodied settings through vision-langu…
View article: RLinf: Flexible and Efficient Large-scale Reinforcement Learning via Macro-to-Micro Flow Transformation
RLinf: Flexible and Efficient Large-scale Reinforcement Learning via Macro-to-Micro Flow Transformation Open
Reinforcement learning (RL) has demonstrated immense potential in advancing artificial general intelligence, agentic intelligence, and embodied intelligence. However, the inherent heterogeneity and dynamicity of RL workflows often lead to …
View article: Risk Transfer Measures of Shadow Banks: A Case Study of Lufax
Risk Transfer Measures of Shadow Banks: A Case Study of Lufax Open
Lufaxs digital platform lowers barriers to entry for financial services, especially for SMEs and individuals who may be underserved by traditional banks. This contributes to broader financial inclusion, stimulating economic activity by ena…
View article: Mechanistic insights into the effects of Tris-2-butoxyethyl phosphate on multiple cancers using network toxicology and molecular docking
Mechanistic insights into the effects of Tris-2-butoxyethyl phosphate on multiple cancers using network toxicology and molecular docking Open
This study investigates the potential mechanisms by which Tris(2-butoxyethyl) phosphate (TBEP) may contribute to the development and progression of human malignancies. Fourteen representative cancer types were selected to explore the molec…
View article: Associations between Life’s Crucial 9 and severity, all-cause mortality, and cardiovascular mortality in individuals with cardiovascular-kidney-metabolic syndrome: the mediating role of phenotypic age acceleration
Associations between Life’s Crucial 9 and severity, all-cause mortality, and cardiovascular mortality in individuals with cardiovascular-kidney-metabolic syndrome: the mediating role of phenotypic age acceleration Open
Objective The American Heart Association (AHA) recently introduced the concept of cardiovascular-kidney-metabolic (CKM) syndrome. This study aimed to explore the associations of Life’s Crucial 9 (LC9), the recently updated lifestyle guidel…
View article: Piano: A Multi-Constraint Pin Assignment-Aware Floorplanner
Piano: A Multi-Constraint Pin Assignment-Aware Floorplanner Open
Floorplanning is a critical step in VLSI physical design, increasingly complicated by modern constraints such as fixed-outline requirements, whitespace removal, and the presence of pre-placed modules. In addition, the assignment of pins on…
View article: Towards Efficient Multi-Scale Deformable Attention on NPU
Towards Efficient Multi-Scale Deformable Attention on NPU Open
Multi-scale deformable attention (MSDA) is a flexible and powerful feature extraction mechanism for visual tasks, but its random-access grid sampling strategy poses significant optimization challenges, especially on domain-specific acceler…
View article: Re‐examining investor sentiment and stock returns: A replication and extension of Baker and Wurgler (2006)
Re‐examining investor sentiment and stock returns: A replication and extension of Baker and Wurgler (2006) Open
This study replicates and extends Baker and Wurgler's (2006) analysis on investor sentiment's impact on stock returns. We confirm their findings by demonstrating the significant cross‐sectional effect of sentiment in both their original sa…
View article: Hierarchical Multi-Stage BERT Fusion Framework with Dual Attention for Enhanced Cyberbullying Detection in Social Media
Hierarchical Multi-Stage BERT Fusion Framework with Dual Attention for Enhanced Cyberbullying Detection in Social Media Open
Detecting and classifying cyberbullying in social media is hard because of the complex nature of online language and the changing nature of content. This study presents a multi-stage BERT fusion framework. It uses hierarchical embeddings, …
View article: Research on the Influencing Factors of Subjective Well-Being in 'Citywalk' Tourism Contexts: Evidence from Nanjing in China
Research on the Influencing Factors of Subjective Well-Being in 'Citywalk' Tourism Contexts: Evidence from Nanjing in China Open
This research aims to investigates the factors influencing subjective well-being in Citywalk tourism. Using structural equation modelling, we found that the destination has a positive impact on tourists' well-being, which is conveyed throu…
View article: Attention-Mechanism-Based Neural Latent-Factorization-of-Tensors Model
Attention-Mechanism-Based Neural Latent-Factorization-of-Tensors Model Open
High-Dimensional and Incomplete (HDI) tensors contain a wealth of knowledge and patterns, which are typically utilized to characterize complex relationships between entities in a variety of industrial applications. Currently, the neural ne…
View article: Idiosyncrasies in Large Language Models
Idiosyncrasies in Large Language Models Open
In this work, we unveil and study idiosyncrasies in Large Language Models (LLMs) -- unique patterns in their outputs that can be used to distinguish the models. To do so, we consider a simple classification task: given a particular text ou…
View article: ClinKD: Cross-Modal Clinical Knowledge Distiller For Multi-Task Medical Images
ClinKD: Cross-Modal Clinical Knowledge Distiller For Multi-Task Medical Images Open
Medical Visual Question Answering (Med-VQA) represents a critical and challenging subtask within the general VQA domain. Despite significant advancements in general VQA, multimodal large language models (MLLMs) still exhibit substantial li…
View article: A multi-dimensional decision framework based on the XGBoost algorithm and the constrained parametric approach
A multi-dimensional decision framework based on the XGBoost algorithm and the constrained parametric approach Open
With the wide access to data and advanced technologies, organizations and firms prefer to use data-based and interpretable analytics to deal with uncertain and cognitive decision-making problems. In this regard, this study considers quanti…
View article: SCCD: A Session-based Dataset for Chinese Cyberbullying Detection
SCCD: A Session-based Dataset for Chinese Cyberbullying Detection Open
The rampant spread of cyberbullying content poses a growing threat to societal well-being. However, research on cyberbullying detection in Chinese remains underdeveloped, primarily due to the lack of comprehensive and reliable datasets. No…
View article: A Hybrid Attention Framework for Fake News Detection with Large Language Models
A Hybrid Attention Framework for Fake News Detection with Large Language Models Open
With the rapid growth of online information, the spread of fake news has become a serious social challenge. In this study, we propose a novel detection framework based on Large Language Models (LLMs) to identify and classify fake news by i…
View article: Enhancing User Intent for Recommendation Systems via Large Language Models
Enhancing User Intent for Recommendation Systems via Large Language Models Open
Recommendation systems play a critical role in enhancing user experience and engagement in various online platforms. Traditional methods, such as Collaborative Filtering (CF) and Content-Based Filtering (CBF), rely heavily on past user int…
View article: Enhancing User Intent for Recommendation Systems via Large Language Models
Enhancing User Intent for Recommendation Systems via Large Language Models Open
Recommendation systems play a critical role in enhancing user experience and engagement in various online platforms. Traditional methods, such as Collaborative Filtering (CF) and Content-Based Filtering (CBF), rely heavily on past user int…
View article: The Application of Large Language Models in Recommendation Systems
The Application of Large Language Models in Recommendation Systems Open
The integration of Large Language Models into recommendation frameworks presents key advantages for personalization and adaptability of experiences to the users. Classic methods of recommendations, such as collaborative filtering and conte…
View article: Adapting to uncertainty: A quantitative investment decision model with investor sentiment and attention analysis
Adapting to uncertainty: A quantitative investment decision model with investor sentiment and attention analysis Open
In the face of global uncertainties, including pandemics, economic fluctuations, disruptions in supply chains, major disasters, wars, and impending economic crises, the financial landscape and the impact of investor sentiment on the return…
View article: Novel Multiple Criteria Group Decision-Making Method Based on Hesitant Fuzzy Clustering Algorithm
Novel Multiple Criteria Group Decision-Making Method Based on Hesitant Fuzzy Clustering Algorithm Open
The aim of this paper is to investigate the multiple criteria group decision-making problem in which the evaluation values provided by experts construct some hesitant fuzzy numbers. We first analyze the characteristics of hesitant fuzzy nu…
View article: Overcoming barriers and seizing opportunities in the innovative adoption of next-generation digital technologies
Overcoming barriers and seizing opportunities in the innovative adoption of next-generation digital technologies Open
In a world increasingly driven by digital transformation, emerging technologies such as blockchain, artificial intelligence, and the metaverse hold great promise for economic and social progress. However, their widespread adoption is influ…