Abstract: Learning reward models from pairwise comparisons is a fundamental component in a number of domains, in-cluding autonomous control, conversational agents, and rec-ommendation systems, as part ...

Abstract: Ocean color remote sensing has advanced over the past 40 years, with algorithms primarily rooted in radiative transfer theory to estimate various ocean color properties. To address the ...