Evidence receipt / belief
Published · transcript-backedJeff Dean: belief
12 Feb 2025 Dwarkesh Podcast Jeff Dean & Noam Shazeer — 25 years at Google: from PageRank to AGI
“You don't necessarily record the actual gradient update in a log or something, but you could replay that log of operations so that you get repeatability. Then I think you'd be happy.”
Source trail
Everything needed to verify it.
- Speaker
- Jeff Dean
- Attribution
- Verified speaker
- Claim type
- belief
- Recorded
- 12 Feb 2025
- Publisher
- Dwarkesh Podcast
Transcript context
…Ooh. So, one way to do that is you effectively record the sequence of operations, like which gradient update happened and when and on which batch of data. You don't necessarily record the actual gradient update in a log or something, but you could replay that log of operations so that you get repeatability. Then I think you'd be happy. Possibly. At least you could debug what happened, but you wouldn't be able to necessarily compare two training runs. Because, okay, I made one change in the hyperparameter, but also I had a-…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.