PADA: Example-based Prompt Learning for on-the-fly Adaptation to Unseen Domains

TL;DR

PADA introduces an example-based prompt learning method that enables on-the-fly adaptation to unseen domains in NLP tasks without prior target domain data, significantly improving performance over baselines.

Contribution

The paper proposes PADA, a novel prompt learning approach that generates domain-specific prompts for unseen domains, facilitating effective zero-shot domain adaptation in NLP.

Findings

PADA outperforms strong baselines in 14 multi-source adaptation scenarios.

It effectively generates domain-related prompts that improve task performance.

The method works across text classification and sequence tagging tasks.

Abstract

Natural Language Processing algorithms have made incredible progress, but they still struggle when applied to out-of-distribution examples. We address a challenging and underexplored version of this domain adaptation problem, where an algorithm is trained on several source domains, and then applied to examples from unseen domains that are unknown at training time. Particularly, no examples, labeled or unlabeled, or any other knowledge about the target domain are available to the algorithm at training time. We present PADA: An example-based autoregressive Prompt learning algorithm for on-the-fly Any-Domain Adaptation, based on the T5 language model. Given a test example, PADA first generates a unique prompt for it and then, conditioned on this prompt, labels the example with respect to the NLP prediction task. PADA is trained to generate a prompt which is a token sequence of unrestricted length, consisting of Domain Related Features (DRFs) that characterize each of the source domains. Intuitively, the generated prompt is a unique signature that maps the test example to a semantic space spanned by the source domains. In experiments with 3 tasks (text classification and sequence tagging), for a total of 14 multi-source adaptation scenarios, PADA substantially outperforms strong baselines.