Zeqiu Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Zeqiu Xu has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Peiyang Yu
Carnegie Mellon University
5 shared publications - Jiani Wang
Stanford University, USA
4 shared publications - Xiaochuan Xu
Carnegie Mellon University, USA
4 shared publications - Jingyuan Yi
Carnegie Mellon University
2 shared publications - Tianyi Huang
University of California, Berkeley, USA
2 shared publications
Latest Publications
- Open Access
- 27 March 2025
- Published by MDPI AG in Preprints.org
A Survey of Reinforcement Learning-Driven Knowledge Distillation: Techniques, Challenges, and Applications
As deep learning models are widely applied across various domains, a critical challenge is how to compress models while maintaining high reasoning capability. Knowledge distillation, an effective technique for model compression, has been used to enhance the performance of lightweight models. However, traditional distillation methods are limited when dealing with complex reasoning tasks. Reinforcement learning (RL) offers a novel approach to knowledge distillation by optimizing the reasoning strategies of teacher models, generating more efficient decision paths, and providing more valuable learning content for student models. This paper reviews the latest advancements in combining reinforcement learning with knowledge distillation, focusing on policy distillation, value function distillation, and dynamic reward-guided distillation methods. It also discusses the challenges faced by RL-driven distillation, such as simplifying complex strategies, addressing temporal dependencies, and balancing exploration and exploitation, and suggests possible solutions. Finally, this paper explores the applications of RL-driven knowledge distillation in fields such as game AI, robotic control, and dialogue systems, and outlines future research directions, including automated distillation, multimodal distillation, and challenges in federated learning.
- 27 March 2025
- Published by MDPI AG in Preprints.org
Open AccessA Survey of Reinforcement Learning-Driven Knowledge Distillation: Techniques, Challenges, and Applications
As deep learning models are widely applied across various domains, a critical challenge is how to compress models while maintaining high reasoning capability. Knowledge distillation, an effective technique for model compression, has been used to enhance the performance of lightweight models. However, traditional distillation methods are limited when dealing with complex reasoning tasks. Reinforcement learning (RL) offers a novel approach to knowledge distillation by optimizing the reasoning strategies of teacher models, generating more efficient decision paths, and providing more valuable learning content for student models. This paper reviews the latest advancements in combining reinforcement learning with knowledge distillation, focusing on policy distillation, value function distillation, and dynamic reward-guided distillation methods. It also discusses the challenges faced by RL-driven distillation, such as simplifying complex strategies, addressing temporal dependencies, and balancing exploration and exploitation, and suggests possible solutions. Finally, this paper explores the applications of RL-driven knowledge distillation in fields such as game AI, robotic control, and dialogue systems, and outlines future research directions, including automated distillation, multimodal distillation, and challenges in federated learning.
- Open Access
Hierarchical Multi-Stage BERT Fusion Framework with Dual Attention for Enhanced Cyberbullying Detection in Social Media
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, dual attention mechanisms, and extra features to improve detection of cyberbullying content. The framework combines BERT embeddings with features like sentiment and topic information. It uses self-attention and cross-attention to align features and has a hierarchical classification head for multi-category classification. A dynamic loss balancing strategy helps optimize learning and improves accuracy, precision, recall, and F1-score. These results show the model's strong performance and potential for broader use in analyzing social media content.
Open AccessHierarchical Multi-Stage BERT Fusion Framework with Dual Attention for Enhanced Cyberbullying Detection in Social Media
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, dual attention mechanisms, and extra features to improve detection of cyberbullying content. The framework combines BERT embeddings with features like sentiment and topic information. It uses self-attention and cross-attention to align features and has a hierarchical classification head for multi-category classification. A dynamic loss balancing strategy helps optimize learning and improves accuracy, precision, recall, and F1-score. These results show the model's strong performance and potential for broader use in analyzing social media content.
- Open Access
- 3 February 2025
- Published by MDPI AG in Preprints.org
Challenges and Innovations in LLM-Powered Fake News Detection: A Synthesis of Approaches and Future Directions
The pervasiveness of the dissemination of fake news through social media platforms poses critical risks to the trust of the general public, societal stability, and democratic institutions. This challenge calls for novel methodologies in detection, which can keep pace with the dynamic and multi-modal nature of misinformation. Recent works include powering the detection using large language model advances in multimodal frameworks, methodologies using graphs, and adversarial training in the literature of fake news. Based on the different approaches which can bring success, some key highlights will be underlined: enhanced LLM-improves accuracy through more advanced semantics and cross-modality fusion for robust detections. The review further identifies critical gaps in adaptability to dynamic social media trends, real-time, and cross-platform detection capabilities. Future directions underline the development of style-agnostic models, cross-lingual detection frameworks, and robust policies with a view to mitigating LLM-driven misinformation. This synthesis thus lays a concrete foundation for those researchers and practitioners committed to reinforcing fake news detection systems with complications that keep on growing in the digital landscape.
- 3 February 2025
- Published by MDPI AG in Preprints.org
Open AccessChallenges and Innovations in LLM-Powered Fake News Detection: A Synthesis of Approaches and Future Directions
The pervasiveness of the dissemination of fake news through social media platforms poses critical risks to the trust of the general public, societal stability, and democratic institutions. This challenge calls for novel methodologies in detection, which can keep pace with the dynamic and multi-modal nature of misinformation. Recent works include powering the detection using large language model advances in multimodal frameworks, methodologies using graphs, and adversarial training in the literature of fake news. Based on the different approaches which can bring success, some key highlights will be underlined: enhanced LLM-improves accuracy through more advanced semantics and cross-modality fusion for robust detections. The review further identifies critical gaps in adaptability to dynamic social media trends, real-time, and cross-platform detection capabilities. Future directions underline the development of style-agnostic models, cross-lingual detection frameworks, and robust policies with a view to mitigating LLM-driven misinformation. This synthesis thus lays a concrete foundation for those researchers and practitioners committed to reinforcing fake news detection systems with complications that keep on growing in the digital landscape.
- Open Access
A Hybrid Attention Framework for Fake News Detection with Large Language Models
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 integrating textual statistical features and deep semantic features. Our approach utilizes the contextual understanding capability of the large language model for text analysis and introduces a hybrid attention mechanism to focus on feature combinations that are particularly important for fake news identification. Extensive experiments on the WELFake news dataset show that our model significantly outperforms existing methods, with a 1.5\% improvement in F1 score. In addition, we assess the interpretability of the model through attention heat maps and SHAP values, providing actionable insights for content review strategies. Our framework provides a scalable and efficient solution to deal with the spread of fake news and helps build a more reliable online information ecosystem.
Open AccessA Hybrid Attention Framework for Fake News Detection with Large Language Models
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 integrating textual statistical features and deep semantic features. Our approach utilizes the contextual understanding capability of the large language model for text analysis and introduces a hybrid attention mechanism to focus on feature combinations that are particularly important for fake news identification. Extensive experiments on the WELFake news dataset show that our model significantly outperforms existing methods, with a 1.5\% improvement in F1 score. In addition, we assess the interpretability of the model through attention heat maps and SHAP values, providing actionable insights for content review strategies. Our framework provides a scalable and efficient solution to deal with the spread of fake news and helps build a more reliable online information ecosystem.
- Open Access
- 8 January 2025
- Published by MDPI AG in Preprints.org
Enhancing User Intent for Recommendation Systems via Large Language Models
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 interactions or item features. However, these models often fail to capture the dynamic and evolving nature of user preferences. To address these limitations, we propose DUIP (Dynamic User Intent Prediction), a novel framework that combines LSTM networks with Large Language Models (LLMs) to dynam- ically capture user intent and generate personalized item recommendations. The LSTM component models the sequential and temporal dependencies of user behavior, while the LLM utilizes the LSTM-generated prompts to predict the next item of interest. Experimental results on three diverse datasets—ML-1M, Games, and Bun- dle—show that DUIP outperforms a wide range of baseline models, demonstrating its ability to handle the cold-start problem and real-time intent adaptation. The integration of dynamic prompts based on recent user interactions allows DUIP to provide more accurate, context-aware, and personalized recommendations. Our findings suggest that DUIP is a promising approach for next-generation recommendation systems, with potential for further improvements in cross-modal recommendations and scalability.
- 8 January 2025
- Published by MDPI AG in Preprints.org
Open AccessEnhancing User Intent for Recommendation Systems via Large Language Models
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 interactions or item features. However, these models often fail to capture the dynamic and evolving nature of user preferences. To address these limitations, we propose DUIP (Dynamic User Intent Prediction), a novel framework that combines LSTM networks with Large Language Models (LLMs) to dynam- ically capture user intent and generate personalized item recommendations. The LSTM component models the sequential and temporal dependencies of user behavior, while the LLM utilizes the LSTM-generated prompts to predict the next item of interest. Experimental results on three diverse datasets—ML-1M, Games, and Bun- dle—show that DUIP outperforms a wide range of baseline models, demonstrating its ability to handle the cold-start problem and real-time intent adaptation. The integration of dynamic prompts based on recent user interactions allows DUIP to provide more accurate, context-aware, and personalized recommendations. Our findings suggest that DUIP is a promising approach for next-generation recommendation systems, with potential for further improvements in cross-modal recommendations and scalability.