Rashid: A Cipher-Based Framework for Exploring In-Context Language Learning

TL;DR

The paper introduces Rashid, a framework that uses reversible ciphering of high-resource languages to create truly unseen languages, enabling comprehensive study and evaluation of in-context learning across diverse languages and tasks.

Contribution

Rashid provides a novel method to simulate unseen languages from high-resource ones, facilitating large-scale, resource-efficient exploration of in-context language learning phenomena.

Findings

Rashid enables systematic evaluation of ICLL methods across multiple languages.

Utilizing resources can improve ICLL performance.

The framework reveals insights into current ICLL capabilities and limitations.

Abstract

Where there is growing interest in in-context language learning (ICLL) for unseen languages with large language models, such languages usually suffer from the lack of NLP tools, data resources, and researcher expertise. This means that progress is difficult to assess, the field does not allow for cheap large-scale experimentation, and findings on ICLL are often limited to very few languages and tasks. In light of such limitations, we introduce a framework (Rashid), for studying ICLL wherein we reversibly cipher high-resource languages (HRLs) to construct truly unseen languages with access to a wide range of resources available for HRLs, unlocking previously impossible exploration of ICLL phenomena. We use our framework to assess current methods in the field with SOTA evaluation tools and manual analysis, explore the utility of potentially expensive resources in improving ICLL, and test ICLL strategies on rich downstream tasks beyond machine translation. These lines of exploration showcase the possibilities enabled by our framework, as well as providing actionable insights regarding current performance and future directions in ICLL.