arXiv 2609.20800
JEPA-Anything: Learning Predictive Models across Different Worlds
By Taoyong Cui, Zhongyao Wang, et al.
Published 2026-09-17
Discussion
Read the public discussion and references gathered around this paper.
World modeling enables intelligence to anticipate consequences, guide interventions, and learn from interaction. Yet predictive models remain domain-specific: can a common learning principle support world modeling across radically different systems? We introduce JEPA-Anything, a domain-agnostic framework based on orthogonal predictive factorization (OPF). Extending joint-embedding predictive architectures, OPF decom…