Tian-Yi Chen

Learning to Follow In-Context Watermark Instructions via Self-Distillation

This work proposes a self-contained two-stage training method, requiring no distillation from a stronger model, no manual annotation, and no pre-existing ICW IF ability, and introduces $\mathsf{ICWBench}$, a benchmark of three verifiable ICW instruction families, each scored on both detectability and answer quality.