DeepSeek-R1-Zero, every mention
3 scenes · ← back to DeepSeek-R1-Zero
tap a year for its mentions
every year anyone Marco Mascorro 12Guido Appenzeller 5
Verbatim, from the transcripts: the passages where DeepSeek-R1-Zero comes up
DeepSeek, Reasoning Models, and the Future of LLMs
- ▶ 7:15 Marco Mascorro In between, there was this model that was called R one zero, which was basically, they took a V three base model, 4 times in the scene
- ▶ 9:55 Guido Appenzeller So the first thing they did was basically they trained this thing called Deep Seek R one zero, which is great at reasoning, but has a lot of other problems. 5 times in the scene
- ▶ 13:02 Marco Mascorro So, basically, they applied this pipeline, and they saw these few things that were interesting in R one zero, right? 8 times in the scene