The recording of the GAN test of time talk by @dwf is now publicly available: https://
neurips.cc/virtual/2024/t
est-of-time/105032
…
RESEARCH
-
GAN Test of Time Talk Recording Now Available
By
–
-

Nature publishes evolutionary model merging optimization research
By
–
論文「Evolutionary Optimization of Model Merging Recipes」が論文誌「Nature Machine Intelligence」に採択され本日掲載されました。最新バージョンでは本アプローチをさらに実証する新たな実験結果を含んでいます。ぜひ以下からご覧ください。 https://
nature.com/articles/s4225
6-024-00975-8
… Sakana -

Nature Publishes Evolutionary Optimization Model Merging Paper
By
–
We’re pleased to announce that our paper, “Evolutionary Optimization of Model Merging Recipes,” has been accepted to Nature Machine Intelligence and published today! The latest version includes new experimental results that further validate our approach. https://
nature.com/articles/s4225
6-024-00975-8
… -
Cade Metz’s New York Times explainer on DeepSeek and AI
By
–
For everyone not living and breathing AI tech because they've got other things going on in their lives, we found @CadeMetz's piece in the Times to be the clearest explainer about what's happening and why. Nice work. nytimes.com/2025/01/27/techn…
-
GRPO motivation: computational and memory efficiency clarification
By
–
Awesome, thanks a lot for this direct head-to-head comparison! Actually, after reading the R1 paper, wasn't the pure motivation behind GRPO computational (/memory) efficiency?
-
Google Fired Me for Questioning Model Size Strategy
By
–
At Google, I asked why they were fixated on building THE LARGEST model. Why are you going for size? What function are you trying to achieve? Why is the thing you were upset about that you didn't have THE LARGEST model? They responded by firing me
-
Media Propaganda Around OpenAI DeepMind Exposed
By
–
No one & I mean NO ONE wanted to listen to us when we screamed at the top of our lungs that the emperor has no clothes… Nothing else has taught me more about how propaganda works as seeing how everyone treated OpenAI, Deepmind & friends since their inception.
-
DeepSeek R1 True Training Costs Beyond $5.567M
By
–
The R1 model wasn't free, and the DeepSeek R1 paper doesn't talk about costs at all. The $5.567M cost for DeepSeek is for when you have all the ingredients in front of you – running that recipe costs $5.567M. It doesn't consider the cost of procuring those ingredients (in this
-
DeepSeek AGI debate raises ethical concerns beyond geopolitics
By
–
No wonder they're losing their shit with the DeepSeek thing. Addendum: whether DeepSeek or OpenAI, there's no such thing as AGI & attempting to build it results in harm: https://
firstmonday.org/ojs/index.php/
fm/article/view/13636
…. Don't make this into a "US vs China US has to win on AGI" thing please -
Algorithms vs Unlimited Resources: OpenAI’s Opposite Approach
By
–
What's interesting about algorithms? The challenge of doing X with as few resources as possible (memory, etc). What do OpenAI + friends do? The opposite. "Lets assume we have unlimited resources & try to build a god."