Training Data Influence Decomposition

STRIDE: Training Data Attribution via Sparse Recovery from Subset Perturbations

Announce Type: new Abstract: Training Data Attribution (TDA) seeks to trace a model's predictions back to its training data. The gold standard for TDA relies on causal interventions, observing how a model changes when data is added or removed, but repeated retraining is computationally challenging for Large Language Models (LLMs). Consequently, most approaches approximate this effect in the parameter space using gradients.