Have you ever been confused between Matt Damon and Mark Wahlberg, Daniel Radcliffe and Elijah Wood, or Margot Robbie and Jaime Pressly? This talk delves into the fascinating world of finding actor look-alikes using multi-modal large language models (LLMs).

By leveraging textual and visual data, these advanced models create detailed embeddings of actors, allowing for the identification of similar-looking pairs. We will explore how these embeddings can uncover clusters of look-alikes within the vast universe of actors and determine the degree of resemblance between different actors, providing insights into the intriguing overlaps in facial features across Hollywood.