Prompts with different control signals (e.g., length, keywords, etc.) can be used to control text summarization. When control signals are available, they can control the properties of generated summaries and potentially improve summarization quality (since more information are given). Unfortunately, control signals are not already available during inference time. In this paper, we propose Lotus (shorthand for Latent Prompt Tuning for Summarization), which is a single model that can be applied in both controlled and uncontrolled (without control signals) modes. During training, Lotus learns latent prompt representations from prompts with gold control signals using a contrastive learning objective. Experiments show Lotus in uncontrolled mode consistently improves upon strong (uncontrollable) summarization models across four different summarization datasets. We also demonstrate generated summaries can be controlled using prompts with user specified control tokens.
翻译:具有不同控制信号( 如长度、 关键字等) 的提示可用于控制文本总和。 当有控制信号时, 它们可以控制生成摘要的属性, 并有可能提高总和质量( 因为提供了更多信息 ) 。 不幸的是, 在推断时间里, 控制信号已经不存在。 在本文中, 我们提出Lotus( 用于Litetent Express Tutning for Summarization), 这是一种单一的模式, 可以同时用于控制和不受控制( 没有控制信号 ) 模式。 在培训期间, Lotus 利用对比性学习目标, 从带有金控制信号的提示中学习潜在的提示。 实验显示, 不受控制的Lotus 在四个不同总和数据集的强( 无法控制的) 总和模型下, 持续改善不受控制的Lotus 模式。 我们还表明, 生成的摘要可以用用户指定控制符号的提示来控制 。