dictionary-learning

This paper introduces a concept-based interpretability framework for end-to-end autonomous driving models, using unsupervised dictionary learning via sparse autoencoders to decompose driving behavior into human-interpretable concepts. The framework enables analysis and targeted correction of model decisions, improving overall driving performance.

This paper proposes replacing the inner product scoring in sparse autoencoders with a learned combination of cosine similarity and input magnitude, showing that the resulting features are more interpretable and concept-aligned, with the optimizer consistently preferring cosine over inner product.

The paper introduces diverse dictionary learning, showing that key set-theoretic relationships among latent variables can be identified from observational data without strong assumptions, enabling partial or full identifiability with minimal inductive bias.