May 21, 2023 front

ML Newsnew | past | best | rss | submit

Stories from May 21, 2023 (UTC)

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

1.	GPT detectors are biased against non-native English writers (arxiv.org)
	338 points by giuliomagnifico on May 21, 2023 \| hide \| past \| pdf \| 274 comments
2.	DarkBERT: A Language Model for the Dark Side of the Internet (arxiv.org)
	142 points by rajtilakjee on May 21, 2023 \| hide \| past \| pdf \| 59 comments
3.	In Defense of Pure 16-Bit Floating-Point Neural Networks (arxiv.org)
	92 points by belter on May 21, 2023 \| hide \| past \| pdf \| 45 comments
4.	Perfectly secure steganography using minimum entropy coupling (arxiv.org)
	71 points by Pseudomanifold on May 21, 2023 \| hide \| past \| pdf \| 21 comments
5.	Neural Network Architecture Beyond Width and Depth (arxiv.org)
	70 points by StrauXX on May 21, 2023 \| hide \| past \| pdf \| 11 comments
6.	Towards Expert-Level Medical Question Answering with Large Language Models (arxiv.org)
	3 points by tontonius on May 21, 2023 \| hide \| past \| pdf \| discuss
7.	Invisible Backdoor Attacks in the Frequency Domain Against Deep Neural Networks (arxiv.org)
	2 points by belter on May 21, 2023 \| hide \| past \| pdf \| 1 comment
8.	Visual Question Answering: Techniques and Common Trends in Recent Literature (arxiv.org)
	2 points by PaulHoule on May 21, 2023 \| hide \| past \| pdf \| discuss
9.	Evidence of Meaning in Language Models Trained on Programs (arxiv.org)
	1 point by belter on May 21, 2023 \| hide \| past \| pdf \| 1 comment