33 min listen
Representation Engineering (Activation Hacking)
Representation Engineering (Activation Hacking)
ratings:
Length:
44 minutes
Released:
Feb 28, 2024
Format:
Podcast episode
Description
Recently, we briefly mentioned the concept of “Activation Hacking” in the episode with Karan from Nous Research. In this fully connected episode, Chris and Daniel dive into the details of this model control mechanism, also called “representation engineering”. Of course, they also take time to discuss the new Sora model from OpenAI.
Released:
Feb 28, 2024
Format:
Podcast episode
Titles in the series (100)
OpenAI, reinforcement learning, robots, safety: with Wojciech Zaremba by Practical AI: Machine Learning, Data Science