Lunch Talk - Incremental Slice Finding for ML Model Debugging

SliceLine is a model debugging technique for finding the top-k worst slices (in terms of conjunctions of attributes) where a trained model performs significantly worse than on average. In contrast to other slice finding techniques, SliceLine introduced an intuitive scoring function, effective pruning strategies, and fast linear-algebra-based evaluation strategies. Together, SliceLine is able to find the exact top-K worst slices in the full lattice of possible conjunctions in reasonable time. Recently, we observe an increasing trend towards iterative algorithms that incrementally update the dataset (e.g., selecting samples, augmentation with new instances). Fully computing SliceLine from scratch for every update is unnecessarily wasteful. In this talk, we share our current work on exploring incremental formulations of SliceLine that leverage statistics from previous runs for additional pruning.

Paper