Soft Prompt Tuning for Cross-Lingual Transfer: When Less is More
This work addresses efficient adaptation of pre-trained language models for multilingual tasks, offering a method that reduces computational costs while improving transfer to linguistically distant languages, though it is incremental in its approach.
The paper tackles cross-lingual transfer by using Soft Prompt Tuning with frozen model parameters, showing that this parameter-efficient approach enhances performance for distant languages, with specific improvements in accuracy and F1 scores reported.
Soft Prompt Tuning (SPT) is a parameter-efficient method for adapting pre-trained language models (PLMs) to specific tasks by inserting learnable embeddings, or soft prompts, at the input layer of the PLM, without modifying its parameters. This paper investigates the potential of SPT for cross-lingual transfer. Unlike previous studies on SPT for cross-lingual transfer that often fine-tune both the soft prompt and the model parameters, we adhere to the original intent of SPT by keeping the model parameters frozen and only training the soft prompt. This does not only reduce the computational cost and storage overhead of full-model fine-tuning, but we also demonstrate that this very parameter efficiency intrinsic to SPT can enhance cross-lingual transfer performance to linguistically distant languages. Moreover, we explore how different factors related to the prompt, such as the length or its reparameterization, affect cross-lingual transfer performance.