Chengliang Chai

Not All Documents Are What You Need for Extracting Instruction Tuning Data

Data-efficient Online Training for Direct Alignment in LLMs

CycleIE: Robust Document Information Extraction through Iterative Verification and Refinement

Not All Instances Are Equally Valuable: Towards Influence-Weighted Dataset Distillation

Localized Data Shapley: Accelerating Valuation for Nearest Neighbor Algorithms

Handling Label Noise via Instance-Level Difficulty Modeling and Dynamic Optimization

Harnessing Diversity for Important Data Selection in Pretraining Large Language Models

Retrieval Augmented Imputation using Data Lake Tables