ICML Keynote Proposes "Weight-Space Learning" as New ML Paradigm

Damian Borth's ICML Weight-Space Symmetries workshop keynote proposed treating neural network weights as a new machine learning modality, editable like latents in diffusion models. His slides, now public, systematically survey the emerging field of weight-space programming and learning.

2026-11-05 ~ 2026-11-05 · 2 related posts

• ICML keynote touts Weight Space Learning: 'Stable Diffusion for neural network weights' — burny_tech · 2026-11-05
• ICML Weight-Space Symmetries keynote slides survey weight-space learning — orvieto_antonio · 2026-11-05