V. S.-J. Huang, “TIDE: Temporal-Aware Sparse Autoencoders for Interpretable Diffusion Transformers in Image Generation”, AAAI, vol. 40, no. 1, pp. 435–443, Mar. 2026.