Amil Dravid's banner

Amil Dravid

@_AmilDravid • 1,278 subscribers

PhD @Berkeley_AI

Shorts

Scaling laws describe how loss changes with scale. Do neurons inside models change predictably too? We study vision and language models up to 30B params and find systematic scaling in neuron universality, specialization, and selectivity. Paper+code: 1/n

Scaling laws describe how loss changes with scale. Do neurons inside models change predictably too? We study vision and language models up to 30B params and find systematic scaling in neuron universality, specialization, and selectivity. Paper+code: 1/n

202,663 views

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.

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 views

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!

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 views