Jake Grigsby

model-agnostic meta-learning (1) transformer architecture (1) multi-task learning (1) few-shot learning (1) imitation learning (1) text classification (1) policy gradient (1) curriculum learning (1) data augmentation (1) in-context learning (1) adversarial training (1) continuous control (1) adversarial attack (1) meta-reinforcement learning (1) task distribution (1) sequence model (1) trajectory feedback (1) multi-task adaptation (1) actor-critic method (1) stochastic neural network (1)