📑 New preprint and code: “From Data to Rewards: a Bilevel Optimization Perspective on Maximum Likelihood Estimation”.