Amil Dravid
@_AmilDravid • 1,095 subscribers
PhD @Berkeley_AI
Shorts
The latent space of earlier generative models like GANS can linearly encode concepts of the data. What if the data was model weights? We present weights2weights, a subspace in diffusion weights that behaves as an interpretable latent space over customized diffusion models.
94,226 просмотров
We've release our code and weights for weights2weights. Check out our demo on Hugging Face 🤗 powered by Gradio. Code: Weights: Demo: Thanks Linoy Tsaban(📍🇯🇵) apolinario 🌐 for the collab!
46,611 просмотров