HyperDiffusionFields (HyDiF): Diffusion-Guided Hypernetworks for Learning Implicit Molecular Neural Fields

Sudarshan Babu, Phillip Lo, Xiao Zhang, Aadi Srivastava, Ali Davariashtiyani, Jason Perera, Michael Maire, Aly A Khan

Published: 23 Sept 2025, Last Modified: 26 Sept 2025AI4D3 2025 PosterEveryoneRevisionsBibTeXCC BY 4.0
Keywords: modelling molecules as fields, hypernetworks, diffusion model
TL;DR: HyperDiffusionFields (HyDiF) models molecules as continuous directional fields via hypernetwork‑conditioned neural implicit functions and a diffusion-based generative process, enabling fine-grained inference and structure-aware generation.
Submission Number: 22