Using the generative nature of a language model to generate task-relevant separators has shown competitive results compared to human-curated prompts like "TL;DR". We demonstrate that even randomly chosen tokens from the vocabulary as separators can achieve near-state-of-the-art performance. We analyse this phenomenon in detail using three different random generation strategies, establishing that the language space is rich with potential good separators, regardless of the underlying language model size. These observations challenge the common assumption that an effective prompt should be human-readable or task-relevant. Experimental results show that using random separators leads to an average 16% relative improvement across nine text classification tasks on seven language models, compared to human-curated separators, and is on par with automatic prompt searching methods.

翻译：利用语言模型的生成性质来生成与任务相关的分隔符，已展现出与人工策划提示（如"TL;DR"）相媲美的竞争性结果。我们证明，即使是从词汇表中随机选取的标记作为分隔符，也能达到接近最优的性能。我们通过三种不同的随机生成策略详细分析了这一现象，证实了语言空间中潜藏着大量优质分隔符，且不受底层语言模型规模的限制。这些观察结果挑战了"有效提示必须具有人类可读性或任务相关性"这一普遍假设。实验结果表明，在七个语言模型的九项文本分类任务中，使用随机分隔符相较于人工策划的分隔符平均带来16%的相对提升，其效果与自动提示搜索方法相当。