SOURCE-LINKED INTELLIGENCE

Data Attribution via Sketched Metadifferentiation

arXiv · AI, language, vision and robotics · article · Sep 14, 2026 · UTC

Data attribution seeks to quantify how individual training examples shape a model's predictions and underpins problems including data valuation, machine unlearning, and model interpretability. Despite having a long line of work, computationally scalable methods often struggle to predict the effect of removing training data in neural networks due to their non-convex nature. To overcome this challenge, metagradient-based methods such as MAGIC (Ilyas and Engstrom, 2025) differentiate each prediction through the entire training run and compute its exact influence with respect to the training data,

Read original source ↗ Open in workspace

recordType

paper

region

Global