Yong Liu

12 papers Tsinghua University, Tsinghua University

• AdaVideoRAG: Omni-Contextual Adaptive Retrieval-Augmented Efficient Long Video Understanding
• Benchmarking Retrieval-Augmented Multimomal Generation for Document Question Answering
• Can LLMs Outshine Conventional Recommenders? A Comparative Evaluation
• Demystifying Reasoning Dynamics with Mutual Information: Thinking Tokens are Information Peaks in LLM Reasoning
• DreamLight: Towards Harmonious and Consistent Image Relighting
• OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain
• P-Law: Predicting Quantitative Scaling Law with Entropy Guidance in Large Recommendation Models
• SSTAG: Structure-Aware Self-Supervised Learning Method for Text-Attributed Graphs
• Sparse MeZO: Less Parameters for Better Performance in Zeroth-Order LLM Fine-Tuning
• Stability and Sharper Risk Bounds with Convergence Rate $\tilde{O}(1/n^2)$
• UltraVideo: High-Quality UHD Video Dataset with Comprehensive Captions
• X-Scene: Large-Scale Driving Scene Generation with High Fidelity and Flexible Controllability