Dear Reviewers:
    This is the supplementary material for our paper titled "DiCache: Let Diffusion Model Determine Its Own Cache".
    We include a demo video that offers a clear performance comparison among DiCache, the non-cached original results, and various existing baselines.
    Additionally, we provide the source code for DiCache, implemented on the video generation model HunyuanVideo and the image generation model Flux.
    We kindly invite you to also consult our appendix (after the bibliography), which contains complementary information that supplements the main paper.
    We hope these materials will provide further insights into our work and assist you in your review process. 
    Thank you very much for your time and efforts! 
Sincerely,
Authors
