Assistant Professor, Carnegie Mellon University
1 paper at NeurIPS 2025
We scale the influence-function-based data valuation method to recent LLMs and their massive training datasets.