“You need to have many layers of evals so that you can check the new model, make sure that it still works the correct way”
♫
Said on Founder Mode
“If you give some of these large frontier models access to the activation layers … they will be able to predict what AlphaFold will do.”
Latent Space: The AI Engineer Podcast · Why AlphaFold Didn't Solve Protein Folding - Pushmeet Kohli, Google DeepMind & Sal Candido, Biohub
Play from 27:11
27:11
Audio streams from the publisher’s own feed. Verbatim does not host or copy it.
What was described
suggested that interpreting AlphaFold2 with a much larger model (accessing activation layers) could reveal how it works
Said nearby on the same subject
“Maybe if there were fewer of those it would be much better and if this forking really happened at points that were interpretable for people”
EB
Eric Bigelow
"The Cognitive Revolution" ·
“I suggest doing no more than three goal details for the quarter… I suggest picking only three that you'll be working on and directing your attention”
LG
Libby Grywe
The Efficient Advisor: Tactical Business Advice for Financial Planners ·
“It could be a Python application that could be running in a server or a container or some serverless application, or it could be inside a very big larger enterprise system.”
SP
Sowjanya Pandruju
Software Engineering Radio - the podcast for professional software developers ·
Moments like this, every Monday
The strongest new moments on your topics, in the speakers' own words. Free for three topics.
No card. No password. One-click unsubscribe.
Track a competitor?
Pro watches up to 25 names or phrases and emails you the morning after they are said.
See Pro, $29 a monthVerbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.