XPrompt: Exploring the Extreme of Prompt Tuning
This addresses the problem of parameter-efficient adaptation for smaller language models, offering an incremental improvement over existing prompt tuning methods.
The paper tackles the performance gap between prompt tuning and fine-tuning for moderate and small-scale language models by identifying that trained prompt tokens can negatively impact tasks. It proposes XPrompt, which uses hierarchical pruning to eliminate negative tokens, achieving competitive performance and closing the gap on SuperGLUE tasks.
Prompt tuning learns soft prompts to condition frozen Pre-trained Language Models (PLMs) for performing downstream tasks in a parameter-efficient manner. While prompt tuning has gradually reached the performance level of fine-tuning as the model scale increases, there is still a large performance gap between prompt tuning and fine-tuning for models of moderate and small scales (typically less than 11B parameters). In this paper, we empirically show that the trained prompt tokens can have a negative impact on a downstream task and thus degrade its performance. To bridge the gap, we propose a novel Prompt tuning model with an eXtremely small scale (XPrompt) under the regime of lottery tickets hypothesis. Specifically, XPrompt eliminates the negative prompt tokens at different granularity levels through a hierarchical structured pruning, yielding a more parameter-efficient prompt yet with a competitive performance. Comprehensive experiments are carried out on SuperGLUE tasks, and the extensive results indicate that XPrompt is able to close the performance gap at smaller model scales.