AI RESEARCH
COMPASS-Hedge: Learning Safely Without Knowing the World
arXiv CS.LG
•
ArXi:2603.22348v1 Announce Type: new Online learning algorithms often faces a fundamental trilemma: balancing regret guarantees between adversarial and stochastic settings and providing baseline safety against a fixed comparator. While existing methods excel in one or two of these regimes, they typically fail to unify all three without sacrificing optimal rates or requiring oracle access to problem-dependent parameters.