Neurocognition, Language and Visual Processing Group Seminar with Dr Marten van Schijndel

We welcome you to our next seminar by the Neurocognition, Language and Visual Processing Group. We will have Dr Marten van Schijndel for a talk. 

• Title: Next Best Bytes Belie Robust Representations
• Abstract:  In language applications, vectors from neural network language models are commonly used to encode semantic information. While these proxies work well for natural language processing tasks, it remains an open question how human-like the resulting representations are. First, I will argue that language model training objectives produce language representations that differ in key ways from human linguistic representations. I will then discuss multiple studies from my group that identify weaknesses in standard tokenization and analysis methods which produce mismatches with human linguistic response patterns, after which I will suggest alternatives to better model human responses.

Join Zoom Meeting: https://Universityofexeter.zoom.us/j/97146951345?pwd=enZGWmljRU5KM2JBeE1aR21kK2xMdz09
Meeting ID: 971 4695 1345
Password: 097263

Dial by your location
        +44 208 080 6591 United Kingdom
        +44 203 481 5240 United Kingdom

For guidance and support on using Zoom, University of Exeter staff and students can visit the University’s Digital Hub.

If you do not have a University of Exeter login please visit Zoom Support.

Short bio: Dr Marten van Schijndel is Assistant Professor of Linguistics at Cornell University. Marten is very interested in the incremental representations that humans use to process language, and in differences between how language is used and how it is processed. To explore these topics, Marten studies the relationships between computational language models and psycholinguistic data (e.g., reading times) and neural network representations of language to understand what aspects of language can be learned from language statistics directly without having experiences in the real world (i.e. through ungrounded learning).

You are also welcome to join the hybrid meeting in Room M1a, Innovation Centre Phase 1 Building by emailing the organiser: idsai@exeter.ac.uk.

Joining our Google group for future seminars and research information: https://groups.google.com/g/neurocognition-language-and-vision-processing-group.

Any queries? Please contact Hang Dong or IDSAI.