Weijie Zhao

Randomized Antipodal Search Done Right for Data Pareto Improvement of LLM Unlearning

UnRe: Zero-Shot LLM Unlearning via Dynamic Contextual Retrieval

DyGB: Dynamic Gradient Boosting Decision Trees with In-Place Updates for Efficient Data Addition and Deletion

VTBench: Evaluating Visual Tokenizer for Autoregressive Image Generation

DMin: Scalable Training Data Influence Estimation for Diffusion Models

A Versioned Unified Graph Index for Dynamic Timestamp-Aware Nearest Neighbor Search

Online Gradient Boosting Decision Tree: In-Place Updates for Adding/Deleting Data

Sparse Watermarking in LLMs with Enhanced Text Quality