ginda307

4 ms·

1. 1.▲
   CAD: Disaggregating Core Attention for Efficient Long-Context LLM Training(hao-ai-lab.github.io)

   6 points by ginda307 10mo ago|0 comments
2. 2.▲
   Disaggregated Inference: 18 Months Later(hao-ai-lab.github.io)

   1 points by ginda307 11mo ago|0 comments
3. 3.▲
   Reasoning Without Hesitating: Efficient Cot Through Certainty Probing(hao-ai-lab.github.io)

   20 points by ginda307 2y ago|5 comments
4. 4.▲
   by ginda307 2y ago

   Just played the game and it seems pretty fun - especially when you see the LLM speaking nonsense haha