Philip S. Yu
YOU?
Author Swipe
View article: A Comprehensive Survey on Multimodal RAG: All Combinations of Modalities as Input and Output
A Comprehensive Survey on Multimodal RAG: All Combinations of Modalities as Input and Output Open
View article: ScaleFormer: Span Representation Cumulation for Long-Context Transformer
ScaleFormer: Span Representation Cumulation for Long-Context Transformer Open
The quadratic complexity of standard self-attention severely limits the application of Transformer-based models to long-context tasks. While efficient Transformer variants exist, they often require architectural changes and costly pre-trai…
View article: Jailbreaking LLMs Through Alignment Vulnerabilities in Out-of-Distribution Settings
Jailbreaking LLMs Through Alignment Vulnerabilities in Out-of-Distribution Settings Open
View article: Frontiers in Graph Machine Learning for the Large Model Era
Frontiers in Graph Machine Learning for the Large Model Era Open
View article: <scp>InterFormer:</scp> Effective Heterogeneous Interaction Learning for Click-Through Rate Prediction
<span>InterFormer:</span> Effective Heterogeneous Interaction Learning for Click-Through Rate Prediction Open
View article: Dialogues Aspect-based Sentiment Quadruple Extraction via Structural Entropy Minimization Partitioning
Dialogues Aspect-based Sentiment Quadruple Extraction via Structural Entropy Minimization Partitioning Open
View article: LLM-Based Human-Agent Collaboration and Interaction Systems: A Survey
LLM-Based Human-Agent Collaboration and Interaction Systems: A Survey Open
View article: The Emerged Security and Privacy of LLM Agent: A Survey with Case Studies
The Emerged Security and Privacy of LLM Agent: A Survey with Case Studies Open
Inspired by the rapid development of Large Language Models (LLMs), LLM agents have evolved to perform complex tasks. LLM agents are now extensively applied across various domains, handling vast amounts of data to interact with humans and e…
View article: The Impact of Digital Transformation on Financing Constraints
The Impact of Digital Transformation on Financing Constraints Open
In the era of the digital economy, the digital transformation of enterprises has a profound impact on their financing capabilities. By comprehensively reviewing relevant studies, this paper clarifies the concepts of digital transformation …
View article: Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models
Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models Open
Message Passing Neural Networks (MPNNs) is the building block of graph foundation models, but fundamentally suffer from oversmoothing and oversquashing. There has recently been a surge of interest in fixing both issues. Existing efforts pr…
View article: Paper2Web: Let's Make Your Paper Alive!
Paper2Web: Let's Make Your Paper Alive! Open
Academic project websites can more effectively disseminate research when they clearly present core content and enable intuitive navigation and interaction. However, current approaches such as direct Large Language Model (LLM) generation, t…
View article: Global-focal Adaptation with Information Separation for Noise-robust Transfer Fault Diagnosis
Global-focal Adaptation with Information Separation for Noise-robust Transfer Fault Diagnosis Open
Existing transfer fault diagnosis methods typically assume either clean data or sufficient domain similarity, which limits their effectiveness in industrial environments where severe noise interference and domain shifts coexist. To address…
View article: RoBCtrl: Attacking GNN-Based Social Bot Detectors via Reinforced Manipulation of Bots Control Interaction
RoBCtrl: Attacking GNN-Based Social Bot Detectors via Reinforced Manipulation of Bots Control Interaction Open
Social networks have become a crucial source of real-time information for individuals. The influence of social bots within these platforms has garnered considerable attention from researchers, leading to the development of numerous detecti…
View article: DeepResearchGuard: Deep Research with Open-Domain Evaluation and Multi-Stage Guardrails for Safety
DeepResearchGuard: Deep Research with Open-Domain Evaluation and Multi-Stage Guardrails for Safety Open
Deep research frameworks have shown promising capabilities in synthesizing comprehensive reports from web sources. While deep research possesses significant potential to address complex issues through planning and research cycles, existing…
View article: Reinforcement Learning from Probabilistic Forecasts for Safe Decision-Making via Conditional Value-at-Risk Planning
Reinforcement Learning from Probabilistic Forecasts for Safe Decision-Making via Conditional Value-at-Risk Planning Open
Sequential decisions in volatile, high-stakes settings require more than maximizing expected return; they require principled uncertainty management. This paper presents the Uncertainty-Aware Markov Decision Process (UAMDP), a unified frame…
View article: AgentDR Dynamic Recommendation with Implicit Item-Item Relations via LLM-based Agents
AgentDR Dynamic Recommendation with Implicit Item-Item Relations via LLM-based Agents Open
Recent agent-based recommendation frameworks aim to simulate user behaviors by incorporating memory mechanisms and prompting strategies, but they struggle with hallucinating non-existent items and full-catalog ranking. Besides, a largely u…
View article: RECODE-H: A Benchmark for Research Code Development with Interactive Human Feedback
RECODE-H: A Benchmark for Research Code Development with Interactive Human Feedback Open
Large language models (LLMs) show the promise in supporting scientific research implementation, yet their ability to generate correct and executable code remains limited. Existing works largely adopt one-shot settings, ignoring the iterati…
View article: Glocal Information Bottleneck for Time Series Imputation
Glocal Information Bottleneck for Time Series Imputation Open
Time Series Imputation (TSI), which aims to recover missing values in temporal data, remains a fundamental challenge due to the complex and often high-rate missingness in real-world scenarios. Existing models typically optimize the point-w…
View article: AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning
AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning Open
LLM-based multi-agent systems excel at planning, tool use, and role coordination, but their openness and interaction complexity also expose them to jailbreak, prompt-injection, and adversarial collaboration. Existing defenses fall into two…
View article: PSG-Agent: Personality-Aware Safety Guardrail for LLM-based Agents
PSG-Agent: Personality-Aware Safety Guardrail for LLM-based Agents Open
Effective guardrails are essential for safely deploying LLM-based agents in critical applications. Despite recent advances, existing guardrails suffer from two fundamental limitations: (i) they apply uniform guardrail policies to all users…
View article: GraphIFE: Rethinking Graph Imbalance Node Classification via Invariant Learning
GraphIFE: Rethinking Graph Imbalance Node Classification via Invariant Learning Open
The class imbalance problem refers to the disproportionate distribution of samples across different classes within a dataset, where the minority classes are significantly underrepresented. This issue is also prevalent in graph-structured d…
View article: Revisiting Multivariate Time Series Forecasting with Missing Values
Revisiting Multivariate Time Series Forecasting with Missing Values Open
Missing values are common in real-world time series, and multivariate time series forecasting with missing values (MTSF-M) has become a crucial area of research for ensuring reliable predictions. To address the challenge of missing data, c…
View article: Advances in Large Language Models for Medicine
Advances in Large Language Models for Medicine Open
Artificial intelligence (AI) technology has advanced rapidly in recent years, with large language models (LLMs) emerging as a significant breakthrough. LLMs are increasingly making an impact across various industries, with the medical fiel…
View article: Utility-based Privacy Preserving Data Mining
Utility-based Privacy Preserving Data Mining Open
With the advent of big data, periodic pattern mining has demonstrated significant value in real-world applications, including smart home systems, healthcare systems, and the medical field. However, advances in network technology have enabl…
View article: Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning
Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning Open
Efficient instruction tuning aims to enhance the ultimate performance of large language models (LLMs) trained on a given instruction dataset. Curriculum learning as a typical data organization strategy has shown preliminary effectiveness i…
View article: Unique Security and Privacy Threats of Large Language Models: A Comprehensive Survey
Unique Security and Privacy Threats of Large Language Models: A Comprehensive Survey Open
With the rapid development of artificial intelligence, large language models (LLMs) have made remarkable advancements in natural language processing. These models are trained on vast datasets to exhibit powerful language understanding and …
View article: MarkDiffusion: An Open-Source Toolkit for Generative Watermarking of Latent Diffusion Models
MarkDiffusion: An Open-Source Toolkit for Generative Watermarking of Latent Diffusion Models Open
We introduce MarkDiffusion, an open-source Python toolkit for generative watermarking of latent diffusion models. It comprises three key components: a unified implementation framework for streamlined watermarking algorithm integrations and…
View article: SGCL: Unifying Self-Supervised and Supervised Learning for Graph Recommendation
SGCL: Unifying Self-Supervised and Supervised Learning for Graph Recommendation Open
View article: ADKGD: Anomaly Detection in Knowledge Graphs with Dual-Channel Training
ADKGD: Anomaly Detection in Knowledge Graphs with Dual-Channel Training Open
In the current development of large language models (LLMs), it is important to ensure the accuracy and reliability of the underlying data sources. LLMs are critical for various applications, but they often suffer from hallucinations and in…
View article: Can Large Language Models Serve as Evaluators for Code Summarization?
Can Large Language Models Serve as Evaluators for Code Summarization? Open