Symmetries and Priors
Tim discusses the absence of symmetries and inductive priors in neural network architecture, questioning their necessity in open-ended systems. Connor and Keith share insights on inductive priors in the CPPN environment encoding, highlighting their implicit presence even when not explicitly stated. Matthew emphasizes the balance between priors and open-endedness, suggesting their role in improving model exploration.In this clip
From this podcast
Machine Learning Street Talk (MLST)
Exploring Open-Ended Algorithms: POET
Related Questions