News for hackers. Stuff that matters.

A comment on OpenAI Furious DeepSeek Might Have Stolen All the Data OpenAI Stole from Us. Parent comment

Re:OpenAI Furious DeepSeek Might Have Stolen All the Data Op... by awestroke on Thursday January 30, 2025 @08:31AM (#42877491)
No. They publish PDFs that hype up their models, but they do not publish anything even resembling a high-level overview of model architecture
[ Parent | on HN ]
Re:OpenAI Furious DeepSeek Might Have Stolen All the Data Op... by jacobgorm on Thursday January 30, 2025 @09:36AM (#42878084)
Given that you can download and use the weights, the model architecture has to be includded as part of that. And I did read a paper from them recently describing their MoE architecture and how it differs from the original GShard.
[ Parent | on HN ]
Re:OpenAI Furious DeepSeek Might Have Stolen All the Data Op... by awestroke on Thursday January 30, 2025 @10:59AM (#42878959)
Excuse me? What weights can you download from OpenAI? gpt2 does not count
[ Parent | on HN ]
Re:OpenAI Furious DeepSeek Might Have Stolen All the Data Op... by jacobgorm on Friday January 31, 2025 @08:01AM (#42887230)
Sorry I meant that DeepSeek release their models. Wrong context.
[ Parent | on HN ]