AI RESEARCH
Split the Differences, Pool the Rest: Provably Efficient Multi-Objective Imitation
arXiv CS.LG
•
ArXi:2605.12000v1 Announce Type: new This work investigates multi-objective imitation learning: the problem of recovering policies that lie on the Pareto front given nstrations from multiple Pareto-optimal experts in a Multi-Objective Marko Decision Process (MOMDP). Standard imitation approaches are ill-equipped for this regime, as naively aggregating conflicting expert trajectories can result in dominated policies. To address this, we