Modern machine learning models may be susceptible to spurious correlations that hold on average but not for the atypical group of samples. To address problem, previous approaches minimize empirical worst-group risk. Despite promise, they often assume each sample belongs one and only group, which does allow expressing uncertainty in labeling. In this paper, we propose a novel framework PG-DRO, e...

For a given target system and apparatus described by quantum theory, the so-called no-programming theorem indicates that family of states called programs in with fixed unitary operation on total distinct dynamics to only if initial are orthogonal each other. The current study aims at revealing whether similar behavior can be observed generalized probabilistic theories (GPTs). Generalizing progr...