October 26, 2024 front

ML Newsnew | past | best | rss | submit

Stories from October 26, 2024 (UTC)

Go back a day, month, or year. Go forward a day.

1.	State-space models can learn in-context by gradient descent (arxiv.org)
	86 points by dsalaj on Oct 26, 2024 \| hide \| past \| pdf \| 58 comments
2.	Moonshine speech to text model 1.7x faster than OpenAI's Whisper, as accurate (arxiv.org)
	2 points by Curiositry on Oct 26, 2024 \| hide \| past \| pdf \| 1 comment
3.	Post-Training Layer Scaling Prevents Forgetting and Enhances Model Merging (arxiv.org)
	1 point by veryluckyxyz on Oct 26, 2024 \| hide \| past \| pdf \| discuss