Sign up to save your podcasts
Or
Share “DeepSeek-r1-0528 Did Not Have a Moment” by Zvi
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
“DeepSeek-r1-0528 Did Not Have a Moment” by Zvi
When r1 was released in January 2025, there was a DeepSeek moment.
When r1-0528 was released in May 2025, there was no moment. Very little talk.
Here is a download link for DeepSeek-R1-0528-GGUF.
It seems like a solid upgrade. If anything, I wonder if we are underreacting, and this illustrates how hard it is getting to evaluate which models are actually good.
What this is not is the proper r2, nor do we have v4. I continue to think that will be a telltale moment.
For now, what we have seems to be (but we’re not sure) a model that is solid for its price and status as an open model, but definitely not at the frontier, that you’d use if and only if you wanted to do something that was a very good fit and played to its strong suits. We likely shouldn’t [...]
---
Outline:
(01:18) We Had a Moment
(05:31) The R2 Moment Will Matter
(08:28) On Your Marks
(16:22) In The News
(18:34) Other Reactions
(27:48) The Distillation Accusation
(30:35) It's Quietly Probably a Solid Model, Sir
---
First published:
June 6th, 2025
Source:
https://www.lesswrong.com/posts/CvxjEkCCq7FpGn5sv/deepseek-r1-0528-did-not-have-a-moment
---
Narrated by TYPE III AUDIO.
---
Images from the article:
Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.
View all episodes
By zviWhen r1 was released in January 2025, there was a DeepSeek moment.
When r1-0528 was released in May 2025, there was no moment. Very little talk.
Here is a download link for DeepSeek-R1-0528-GGUF.
It seems like a solid upgrade. If anything, I wonder if we are underreacting, and this illustrates how hard it is getting to evaluate which models are actually good.
What this is not is the proper r2, nor do we have v4. I continue to think that will be a telltale moment.
For now, what we have seems to be (but we’re not sure) a model that is solid for its price and status as an open model, but definitely not at the frontier, that you’d use if and only if you wanted to do something that was a very good fit and played to its strong suits. We likely shouldn’t [...]
---
Outline:
(01:18) We Had a Moment
(05:31) The R2 Moment Will Matter
(08:28) On Your Marks
(16:22) In The News
(18:34) Other Reactions
(27:48) The Distillation Accusation
(30:35) It's Quietly Probably a Solid Model, Sir
---
First published:
June 6th, 2025
Source:
https://www.lesswrong.com/posts/CvxjEkCCq7FpGn5sv/deepseek-r1-0528-did-not-have-a-moment
---
Narrated by TYPE III AUDIO.
---
Images from the article:
Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.