TL;DR该论文研究了少样本提示模型是否也利用了表象线索，发现在 MNLI、SNLI、HANS 和 COPA 数据集上，尽管模型在表象线索实例上表现良好，但在没有表象线索的实例上的性能往往低于或仅略优于随机准确度。
Abstract
Finetuning large pre-trained language models with a task-specific head has advanced the state-of-the-art on many natural language understanding benchmarks. However, models with a task-specific head require a lot of training data, making them susceptible to learning and exploiting datas