top of page
2023
Project 1
Exploring the "inductive bias of weights" by uncovering inherent structural patterns—such as sparsity, low-rank structures, modularity, equivariance, and symbolic reasoning—that emerge during large-scale foundation model training.
Website Design
Print & Collateral
Editorial Illustration
Client /
J&H
Exploring the "inductive bias of weights" by uncovering inherent structural patterns—such as sparsity, low-rank structures, modularity, equivariance, and symbolic reasoning—that emerge during large-scale foundation model training.
bottom of page