BRAIn: Bayesian Reward-conditioned Amortized Inference for natural language generation from feedback Paper • 2402.02479 • Published Feb 4 • 2
