1 paper
Fabian Schmidt, Markus Enzweiler, Abhinav Valada
Latent world models learn predictive representations for autonomous driving, but the relational semantics these states preserve often remain implicit. We investigate whether traffi…