Dave talking at a meetup with audience in foreground and slides projected behind

A gentle introduction to embeddings at the inaugural GenAI Nework Melbourne meetup

I was thrilled to help kick-off the GenAI Network Melbourne meetup at their first meeting recently. I presented a talk titled Semantic hide and seek – a gentle introduction to embeddings, based on my experiments with Semantle, other representation learning, and some discussion of what it means to use Generative AI in developing new products and services. It was a pleasure to present alongside Rajesh Vasa from A2I2 at Deakin University.

Thanks to Ned, Orian, Scott, Alex, Leonard & co for organising. Looking forward to more fun events in this series!

Check out the slides.

Outline

Background on embeddings

Animated chart mosaic titled This wheelie does not exist. Shows a single dimension (duration) mapping to a 3D latent space which in turn generates a realistic looking 100 sample wheelie trace

The game Semantle and my solvers

  • About the game, and playing with friends
  • Live online solver demo!
  • Solver project aims: experiment with embeddings, automate solutions, explore how people and machines work together on problems
  • Modular solver design and search strategies, illustrated below
Diagram showing modular solver design including semantic model with limited vocabulary, informing search state that determines the next guess to make based on the game state, which can also be influenced by other players' guesses. The cohort and gradient search strategies are shown below.

Reflections on people and machines working together

Diagram Comparing the SECI model with naive automation (means all tacit stages are lost) and augmentation with machines that can help socialise to reinforce the cycle