Looks similar
Were RNNs all we needed?(arxiv.org)
466 points by beefman | 245 comments
arXiv paper proposing minimal LSTM and GRU variants without hidden-state gate dependencies that can be trained in parallel, rivaling recent sequence models.
Best of N Jail Breaking(arxiv.org)
14 points by tzury | 1 comments
Best-of-N Jailbreaking(arxiv.org)
50 points by flyingpumba | 12 comments

![Screenshot of arxiv.org: Serving 70B-scale LLMs efficiently on low-resource edge devices [pdf]](https://hn.urlbox.com/shots/renders/2024/10/3/e613d894-0fb9-482c-8287-3f1b75a65623.jpg)




