DADA: Dialect Adaptation via Dynamic Aggregation of Linguistic Rules
Submission Type: Regular Long Paper
Submission Track: Multilinguality and Linguistic Diversity
Submission Track 2: Human-Centered NLP
Keywords: Dialect Adaptation; Dialect Robustness; Linguistic Diversity; Fairness; Human-Centered NLP
TL;DR: To better address the flexibility of dialects, we propose DADA, a modular approach to imbue SAE-trained models with multi-dialectal robustness from a finer-grained perspective, by composing adapters which handle specific linguistic features.
Abstract: Existing large language models (LLMs) that mainly focus on Standard American English (SAE) often lead to significantly worse performance when being applied to other English dialects. While existing mitigations tackle discrepancies for individual target dialects, they assume access to high-accuracy dialect identification systems. The boundaries between dialects are inherently flexible, making it difficult to categorize language into discrete predefined categories. In this paper, we propose DADA (Dialect Adaptation via Dynamic Aggregation), a modular approach to imbue SAE-trained models with multi-dialectal robustness by composing adapters which handle specific linguistic features. The compositional architecture of DADA allows for both targeted adaptation to specific dialect variants and simultaneous adaptation to various dialects. We show that DADA is effective for both single task and instruction finetuned language models, offering an extensible and interpretable framework for adapting existing LLMs to different English dialects.
Submission Number: 2217
Loading