October 23, 2023 front

ML Newsnew | past | best | rss | submit

Stories from October 23, 2023 (UTC)

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

1.	Embarrassingly Simple Text Watermarks (arxiv.org)
	86 points by PaulHoule on Oct 23, 2023 \| hide \| past \| pdf \| 50 comments
2.	Exploring the Space of Key-Value-Query Models with Intention (arxiv.org)
	3 points by beefman on Oct 23, 2023 \| hide \| past \| pdf \| discuss
3.	Who's Harry Potter? Approximate Unlearning in LLMs [pdf] (arxiv.org)
	3 points by birriel on Oct 23, 2023 \| hide \| past \| pdf \| discuss
4.	Eureka: Human-Level Reward Design via Coding Large Language Models (arxiv.org)
	2 points by rntn on Oct 23, 2023 \| hide \| past \| pdf \| discuss
5.	The Clock and the Pizza: 2 Stories in Mechanistic Explanation of Neural Networks (arxiv.org)
	2 points by andromaton on Oct 23, 2023 \| hide \| past \| pdf \| discuss
6.	Counting and Algorithmic Generalization with Transformers (arxiv.org)
	1 point by PaulHoule on Oct 23, 2023 \| hide \| past \| pdf \| discuss
7.	TacticAI: An AI assistant for football tactics (article by DeepMind) (arxiv.org)
	1 point by falcor84 on Oct 23, 2023 \| hide \| past \| pdf \| discuss
8.	FourCastNet: A Global Data-Driven High-Resolution Weather Model Using AFNOs (arxiv.org)
	1 point by occamschainsaw on Oct 23, 2023 \| hide \| past \| pdf \| discuss