25th November 2020 -11am - Yoav Goldberg: Seminar


TITLE:   Blackbox models are annoying. What can we do about it?


Deep learning models, and in particular the representations resulting from large language models, are black-boxes, making them inherently hard to debug, fix and control. What can we do about this? currently, not much. However, we are starting to tackle this problem. In this talk I will present two of our recent research lines around this problem. In the first one, we move from the continuous states to a symbolic ones (a set of predicted words), which are more debugable and controllable. In the second, we show how to remove or augment certain types of information from the continuous states, to make them more controllable and inspectable. We do this using an iterative projection method. 


Yoav is an associate professor of computer science at Bar Ilan University, and also the research director of AI2 Israel. His research interests include language understanding technologies with real world applications, combining symbolic and neural representations, uncovering latent information in text, syntactic and semantic processing, and interpretability and foundational understanding of deep learning models for text and sequences. He authored a textbook on deep learning techniques for natural language processing, and was among the IEEE's AI Top 10 to Watch in 2018, and a recipient of the Krill Prize in Science in 2017. He received his Ph.D. in Computer Science from Ben Gurion University, and spent time in Google Research as a post-doc.   See https://www.cs.biu.ac.il/~yogo/ for extra information.


Add to your calendar

 vCal  iCal

Nov 25 2020 -

25th November 2020 -11am - Yoav Goldberg: Seminar

This event is co-organised by ILCC and by the UKRI Centre for Doctoral Training in Natural Language Processing, https://nlp-cdt.ac.uk

Blackboard invitation