TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

ArxivGen: Generative Research ArXiv, 100% AI Powered

14 点作者 jeremynixon大约 2 年前

4 条评论

nsthorat大约 2 年前
I cant tell if this is research about generative AI, or AI generated research.<p>The new internet sucks
tysam_and大约 2 年前
If this were truly a blackhat effort, it could improve the quality of empirical papers at conferences specifically by practically forcing a reproducible-by-the-conference-XYZ org of the results on some benchmark using a Docker image, etc etc.<p>On the other hand, for investigative or more purely academic research, this could be another layer of nightmarish crap that people would have to sift through.<p>Also it looks like at least one example is basically cloned from the source set, just with...basically greatly reduced recall of the content (beyond some high points).<p>This feels nightmarish, to be honest. Well, I guess we&#x27;ll find one way to adapt to it, or simply fall back to certain, much more provably-analog methods of doing things....<p>Real: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;pdf&#x2F;1409.6041.pdf" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;pdf&#x2F;1409.6041.pdf</a><p>Fake: <a href="http:&#x2F;&#x2F;arxivgen.com&#x2F;pdfs&#x2F;domain-adaptive_neur-3p5.pdf" rel="nofollow">http:&#x2F;&#x2F;arxivgen.com&#x2F;pdfs&#x2F;domain-adaptive_neur-3p5.pdf</a>
ftxbro大约 2 年前
time to download the internet before it gets filled with whatever this is
评论 #35996735 未加载
marviel大约 2 年前
impressive &amp; terrifying