Generative AI – Intermediate

In this session, we will begin with a Recap of the previous session in May, summarizing key takeaways and setting the stage for the topics ahead. I will talk about the architecture of Neural Networks – how they learn from large volumes of data and delve into techniques to improve Foundation Models (FMs), upon which many downstream applications are built.

Today’s session includes a deep dive into Transfer Learning, which is how knowledge gained from one task can be leveraged for another and Self-Supervised Learning, a methodology that enables models to learn representations without explicit labels.

Next, we will question whether Accuracy is the key consideration, discussing the broader metrics and trade-offs AI practitioners face. We will also address Bias and Hallucinations, highlighting the ethical and practical challenges of deploying AI systems at scale.

Finally, the session concludes with Demos showcasing two notable platforms: