Goal: We simulate Lorenz system let SIREN to represent it, and then we want to see if this representation can be "generalized to the future". Result: No, it cannot.
- To play with the code, we recommend to do this on Google Colab.
- To know more about SIREN: visit their project page: https://vsitzmann.github.io/siren/ or read the paper 'Implicit Neural Representations with Periodic Activation Functions' directly: https://arxiv.org/abs/2006.09661