Implicit Neural Representations of Individual Behavior

Andrew Kang, Priya Narasimhan

We study policy representation learning from unlabeled multi-policy behavioral data. Each episode is generated by a fixed policy, but policy labels are unavailable. This setting ap…