PTP：通过基于扰动的正则化提升提示调优的稳定性与性能
PTP: Boosting Stability and Performance of Prompt Tuning with Perturbation-Based Regularizer
AI总结：
针对现有提示调优训练不稳定的问题，该研究提出基于扰动正则化的PTP算法，通过平滑损失曲面提升稳定性与性能，在SuperGLUE和FewGLUE基准上取得显著提升。
摘要：
近期研究表明，在下游自然语言理解任务上，提示调优比全量微调能更好地发挥大语言模型的能力。然而，现有提示调优方法存在训练不稳定问题：不同随机种子下的得分方差相当大。为解决这一关键问题，我们首先研究发现，可视化后 vanilla 提示调优的损失曲面十分陡峭，输入数据的微小变化就会导致损失曲面的大幅波动，这是引发提示调优不稳定的核心因素。基于这一观察，我们将可平滑损失曲面的基于扰动的正则化引入提示调优，提出了名为基于扰动正则化的提示调优（Prompt Tuning with Perturbation-based regularizer, PTP）的新算法，它不仅能大幅缓解训练不稳定问题，还能提升提示调优的性能。我们设计了两种基于扰动的正则化，包括基于随机噪声的和基于对抗的。特别地，我们提出的扰动在文本空间和嵌入空间都具备灵活性。大量实验证明了所提方法在稳定训练方面的有效性。我们的新算法在 SuperGLUE 和 FewGLUE 基准上分别将当前最优的提示调优方法提升了1.94%和2.34%。
英文摘要：
Recent studies show that prompt tuning can better leverage the power of large language models than fine-tuning on downstream natural language understanding tasks. However, the existing prompt tuning methods have training instability issues, as the variance of scores under different random seeds is quite large. To address this critical problem, we first investigate and find that the loss landscape of vanilla prompt tuning is precipitous when it is visualized, where a slight change of input data can cause a big fluctuation in the loss landscape. This is an essential factor that leads to the instability of prompt tuning. Based on this observation, we introduce perturbation-based regularizers, which can smooth the loss landscape, into prompt tuning. We propose a new algorithm, called Prompt Tuning with Perturbation-based regularizer~(PTP), which can not only alleviate training instability dramatically but also boost the performance of prompt tuning. We design two kinds of perturbation-based regularizers, including random-noise-based and adversarial-based. In particular, our proposed perturbations are flexible on both text space and embedding space. Extensive experiments show the effectiveness of our proposed methods in stabilizing the training. Our new algorithms improve the state-of-the-art prompt tuning methods by 1.94\% and 2.34\% on SuperGLUE and FewGLUE benchmarks, respectively.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
