Dr Aaron Snoswell shares novel research methods at leading international conference on Natural Language Processing
Author Kathy Nickels
Date 20 December 2023
Dr Aaron Snoswell shared research on methods to measure and remove harmful generative capability from language models at the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP), held in Singapore 6-10 December.
The EMNLP Conference stimulated discussions on the ways large language models perform natural language processing (NLP) tasks and applications, and what this means for the future of NLP as a field.
At the conference, Dr Snoswell, from the ARC Centre of Excellence for Automated Decision-Making and Society at QUT, presented work from The Toxicity Scalpel project, which also includes ADM+S members Distinguished Prof Jean Burgess, Prof Nicolas Suzor, Prof Flora Salim, Dr Hao Xue, and Lucinda Nelson.
This project examines how language models used in automated decision-making systems might be improved by making modifications at the pre-trained foundation model stage.
The research was well received at the conference, with generative discussions with other conference attendees, including with researchers from Google and Facebook’s responsible Machine Learning teams.
