ART-AI Seminar
**postponed to the 28th March 13.15pm-14.15pm (BST) **
We are pleased to have Ioana Baldini who is a Research Scientist in the IBM Research AI organization, join us for this ART-AI seminar entitled ‘Auditing Language Models for Bias’. This seminar will take place online on Zoom on Tuesday 28th March 2023, 13.15pm-14.15pm (BST). Joseph Marvin Imperial will be chairing this seminar. For joining instructions, please e-mail [email protected].
Title
Auditing Language Models for Bias
Abstract
The current AI landscape is dominated by larger and larger pre-trained language models (PLMs) applied to different domains without properly understanding how to evaluate their capabilities, and more importantly, how to audit and test their fairness behavior. Auditing PLMs for unwanted social bias is challenging, not only due to their opaque behavior, but also due to the multidisciplinary nature of the work. In the context of PLM-based Natural Language Processing (NLP) systems, this problem is exacerbated by the fact that no single entity controls all the components of an NLP pipeline. As such, it is crucial to assess the fairness of a PLM by only inspecting its outputs. In this talk, we will go over one such example of PLM bias auditing, detailing the challenges and complexities that emerge in the process.
Bio
Ioana Baldini is a Research Scientist in the IBM Research AI organization. As a result of recent collaborations with non-profit organizations on projects under the IBM Science for Social Good Initiative, Ioana has focused her research efforts on natural language processing applied to domains that have potential for social impact. Her current work focuses on different aspects of fairness in natural language processing systems. Her collaboration with Reboot Rx led to the creation and release of public datasets containing information on investigational COVID-19 treatments on cancer outcomes (independent of COVID-19 infection). Previously at IBM Research, Ioana was part of the core research team who developed OpenWhisk, an open source serverless platform, which is now offered as IBM Cloud Functions. In this role, Ioana contributed several components to the OpenWhisk system and helped deploy it as the official IBM serverless platform. Ioana holds a Masters and PhD degrees from the University of Toronto. She received the NSERC (NSF-equivalent in Canada) Canada Graduate Scholarship, the IBM PhD Fellowship, the Canada Google Anita Borg Scholarship, and the IBM Research Division Award (contributions to OpenWhisk).
