2 papers across 2 sessions
We learn a generative model of the Pareto set that can be conditioned on subjective preferences, without retraining, for online multi-objective optimization tasks on discrete/mixed spaces.