Hong-Gee Kim

reinforcement learning (1) continual learning (1) catastrophic forgetting (1) zero-shot learning (1) vision transformer (1) few-shot learning (1) policy optimization (1) sequential decision making (1) text editing (1) grammatical error correction (1) speech recognition (1) prompt learning (1) instruction tuning (1) evaluation metric (1) in-context example (1) multi-head attention (1) masked language modeling (1) entity typing (1) gradient dynamics (1) word error rate (1)