Skip to content
Representation Engineering (Activation Hacking) episode artwork

Feb 28, 2024 43 min

Representation Engineering (Activation Hacking)

from Practical AI Hosted by Daniel Whitenack and Chris Benson

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. Sponsors: Neo4j – Is your code getting dragged down by JOINs and long query times? The problem might be your database…Try simplifying the complex with graphs. Stop asking relational databases to do more than they were made for. Graphs work well for use cases with lots of data connections like supply chain, fraud detection, real-time analytics, and genAI. With Neo4j, you can code in your favorite programming language and against any driver. Plus, it’s easy to integrate into your tech stack. Visit Neo4j.com/developer to get started. Changelog News – A podcast+newsletter combo that’s brief, entertaining & always on-point. Subscribe today. Fly.io – The home of Changelog.com — Deploy your apps and databases close to your users. In minutes you can run your Ruby, Go, Node, Deno, Python, or Elixir app (and databases!) all over the world. No ops required. Learn more at fly.io/changelog and check out the speedrun in their docs. Featuring: Chris Benson – Website, GitHub, LinkedIn, X Daniel Whitenack – Website, GitHub, X Show Notes: Data synthesis for SOTA LLMs with Karan Malhotra from Nous Research (Practical AI #255) Article: Representation Engineering Mistral-7B an Acid Trip OpenAI Sora Upcoming Events: Register for upcoming webinars here!

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

Representation Engineering (Activation Hacking)

0:00 43:34

Chapters

  1. 0:00 Welcome to Practical AI
  2. 0:42 Keeping you fully connected
  3. 1:27 Cool TreeHacks inventions
  4. 5:50 Sponsor: Neo4j
  5. 6:57 Activation hacking & representation engineering
  6. 8:55 Prompt hassles
  7. 10:44 Adding control to prompts
  8. 14:34 Binary control & hidden states
  9. 20:59 Comparing traditional approaches
  10. 23:13 Real world usecase for bias
  11. 24:53 A zoo of behaviors
  12. 27:20 Control vectors in anti-jailbreaking
  13. 28:39 Sponsor: Changelog News
  14. 30:01 Fascinating deep dive
  15. 30:17 OpenAI's Sora
  16. 35:17 Google's Gemma
  17. 39:27 AGI magic
  18. 41:27 Thanks for joining us
  19. 42:47 Outro

No transcript yet.

Requests help us decide what to transcribe next.

Frequently asked questions

How long is this episode of Practical AI?

This episode is 43 minutes long.

When was this Practical AI episode published?

This episode was published on February 28, 2024.

Can I download this Practical AI episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.