Johan Ferret

reinforcement learning (2) transfer learning (1) self-supervised learning (1) reward function (1) credit assignment (1) textual entailment (1) reference-free evaluation (1) parameter-efficient finetuning (1) offline supervised (1) abstractive summarization (1) inference time (1) factual consistency (1) multi-task training (1) temporal ordering (1) multi-objective finetuning (1) reward-based finetuning (1) steerable language model (1) reference-free reward (1) reversibility-aware exploration (1) deep reinforcement learning (1)