Enmao Diao

Graph Tokenization for Bridging Graphs and Transformers

OBCache: Optimal Brain KV Cache Pruning for Efficient Long-Context LLM Inference

RPS: Information Elicitation with Reinforcement Prompt Selection

FedOBP: Federated Optimal Brain Personalization with Few Personalized Parameters

Toward Unifying Group Fairness Evaluation from a Sparsity Perspective

Beyond Expectations: Quantile-Guided Alignment for Risk-Calibrated Language Models

MAP: Multi-Human-Value Alignment Palette

Probe Pruning: Accelerating LLMs through Dynamic Pruning via Model-Probing