Orevaoghene Ahia
PhD Student / Graduate Research Assistant @ University of Washington
reinforcement learning
machine translation
low-resource
modularity
interpretability
language modeling
large language models
self-supervised learning
multilinguality
tokenization
alignment
multlingual
data quality
audit
multilingual data
9
presentations
7
number of views
SHORT BIO
I am a Ph.D. student in Computer Science and Engineering at the University of Washington advised by Noah A. Smith and Yulia Tsvetkov. My research involves topics in Multilingual NLP, Model interpretability, and Model Efficiency and Fairness. I am interested in designing 1.) novel ways to efficiently learn representations fairly across languages with varying linguistic properties. 2.) methods to leverage model interpretability to improve model performance and demote spurious training data artefacts
Presentations
Teaching LLMs to Abstain across Languages via Multilingual Feedback
Shangbin Feng and 8 other authors
Voices Unheard: NLP Resources and Models for Yorùbá Regional Dialects
Orevaoghene Ahia and 7 other authors
Critical Learning Periods: Leveraging Early Training Dynamics for Efficient Data Pruning
Everlyn Chimoto and 5 other authors
MYTE: Morphology-Driven Byte Encoding for Better and Fairer Multilingual Language Modeling
Tomasz Limisiewicz and 4 other authors
Extracting Lexical Features from Dialects via Interpretable Dialect Classifiers
Roy Xie and 3 other authors
Do All Languages Cost the Same? Tokenization in the Era of Commercial Language Models
Orevaoghene Ahia and 6 other authors
Quality at a Glance: An Audit of Web-Crawled Multilingual Datasets
Julia Kreutzer and 51 other authors
The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation
Orevaoghene Ahia and 2 other authors
The Low-Resource Double Bind: An Empirical Study of Pruning for Low-Resource Machine Translation
Orevaoghene Ahia and 2 other authors