Defu Lian

6 papers University of Science and Technology of China

• Accurate KV Cache Eviction via Anchor Direction Projection for Efficient LLM Inference
• Advancing Machine-Generated Text Detection from an Easy to Hard Supervision Perspective
• HawkBench: Investigating Resilience of RAG Methods on Stratified Information-Seeking Tasks
• LogicTree: Improving Complex Reasoning of LLMs via Instantiated Multi-step Synthetic Logical Data
• P-Law: Predicting Quantitative Scaling Law with Entropy Guidance in Large Recommendation Models
• Towards A Generalist Code Embedding Model Based On Massive Data Synthesis