fheinsen | Dark Hacker News

1.	Attention at Constant Cost per Token via Symmetry-Aware Taylor Approximation(arxiv.org) 164 points by fheinsen 149 days ago \| 96 comments
2.	A non-diagonal SSM RNN computed in parallel without requiring stabilization(github.com) 9 points by fheinsen 260 days ago \| 1 comment
3.	1 points by fheinsen 264 days ago \| discuss
4.	Generalized Orders of Magnitude (GOOMs)(github.com) 2 points by fheinsen 265 days ago \| 4 comments
5.	267 days ago \| discuss
6.	3 years ago \| discuss
7.	3 years ago \| discuss