Dongqi Fu

Language in the Flow of Time: Time-Series-Paired Texts Weaved into a Unified Temporal Narrative

Geometric Constraints for Small Language Models to Understand and Expand Scientific Taxonomies

MC-Search: Evaluating and Enhancing Multimodal Agentic Search with Structured Long Reasoning Chains

Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation

OwlEye: Zero-Shot Learner for Cross-Domain Graph Data Anomaly Detection

LinguaMate: Language‑Guided Metamaterial Discovery via Symbolic-Driven Latent Optimization

ReMix: Reinforcement Routing for Mixtures of LoRAs in LLM Finetuning

DAG-MoE: From Simple Mixture to Structural Aggregation in Mixture-of-Experts

Towards Novel Metamaterial Discovery via Latent Space Regulation and Exploration

RAG over Tables: Hierarchical Memory Index, Multi-Stage Retrieval, and Benchmarking

TRQA: Time Series Reasoning Question And Answering Benchmark

SSProNet: Secondary Structure aware Graph Neural Network for Protein Representation Learning

Temporal Heterogeneous Graph Generation with Privacy, Utility, and Efficiency

HeroFilter: Adaptive Spectral Graph Filter for Varying Heterophilic Relations

UniMate: A Unified Model for Mechanical Metamaterial Generation, Property Prediction, and Condition Confirmation

Learning Graph Quantized Tokenizers

Graph4MM: Weaving Multimodal Learning with Structural Information

Learnable Spatial-Temporal Positional Encoding for Link Prediction

Is Attention All You Need for Temporal Link Prediction? A Lightweight Alternative via Learnable Positional Encoding and MLPs

Language Models are Graph Learners

Multi-modal Prompt Learning Empowers Graph Neural Networks with Semantic Knowledge