Tensorflow & Mechanistic Interpretability

The MLAI Meetup is a community for AI researchers and professionals which hosts monthly talks on exciting research. Our format is:

• 6:00 - 6:20: Socializing
• 6:20 - 6:40: Announcements and AI news
• 6:40 - 7:40: Talk(s) and Q&A
• 7:40 - 8:00 Networking
• 8:00: Head to the nearest pub for dinner

Talk 1: Tensorflow

Presenter: Patrick Haralabidis

Abstract: This talk will provide an overview of the advantages of TensorFlow, highlighting how its flexible architecture and comprehensive ecosystem make it a powerful tool for machine learning projects. Attendees will discover the exciting features and applications TensorFlow enables, from deep learning to deploying models in production

Speaker Bio: Patrick is the Engineering Practice Manager - Chapter Lead at Flybuys, with over 15 years of experience in software design, data engineering, and mobile development. He specializes in machine learning and artificial intelligence, having led technical teams at various organizations and contributed technical articles for SitePoint.

Talk 2: Mechanistic Interpretability with Sparse Autoencoders

Presenter: Louka Ewington-Pitsos

Abstract: Wouldn't it be nice if you could see what was going on inside Large Language Models like ChatGPT? Turns out, now you can (kind of)! Thanks to the excellent work of researchers at Anthropic (and others) we are now able to pinpoints exactly what concepts (things like sadness, joy, or Lebron James) feature most prominently at every token of computation. This talk will be an overview of this fascinating new development.

Speaker Bio: Louka has founded multiple start-ups and given talks on Machine Learning at conferences around Australia. He currently works as a Principle Data Scientist at Shavik.AI and tries to contribute to machine learning research in his spare time.