arxiv

The ICLR 2022 Outstanding Paper Award was released, and a work by scholars from Zhejiang University, Tsinghua University and Renmin University was selected
thesis iclr arxiv dpm
04-22
Is it really safe to outsource model training? New research: Outsourcers may implant backdoors to control bank lending
algorithm Classifiers The heart of the machine arxiv
04-21
The large model review research signed by hundreds of scholars was questioned for "plagiarism", and the Zhiyuan Research Institute officially issued an apology letter
Zhiyuan Research Institute nicholas carlini models arxiv
04-14
Zhiyuan apologizes: There are problems with some of the 16 articles, and a third-party independent review has been initiated
Zhiyuan Research Institute nicholas carlini Wisdom Source arxiv Beijing Zhiyuan Artificial Intelligence Research Institute
04-13
Engaged in multimodality and don't know the latest progress? The first review of visual-language pretraining written by the Institute of Automation, Chinese Academy of Sciences
vision Institute of Automation, Chinese Academy of Sciences Pre-training arxiv
03-28
Rare: Well-known physicists are banned and deleted by arXiv, "scientific debates should also talk about civilization"?
arxiv physics thesis University of California, San Diego
03-17
Physicists raved the natural cover room temperature superconductivity paper on arXiv and were banned for 6 months
arxiv Superconducting thesis physics The heart of the machine Preprints
03-15
CVPR 2022! 2067 papers were accepted: early dissemination was really rejected
cvpr The heart of the machine He Kaiming arxiv
03-02
Within 30 days, the emerging compounds were discovered, and the artificial intelligence platform linked up to subvert the research and development of traditional drugs
artificial intelligence Silicon Intelligence Alex Zavolokov arxiv
02-25
Dashen developed the new H5 version of arXiv, saying goodbye to formula typographical errors in one step, and the mobile phone can easily read the literature
arxiv h5 would latex
02-13
When self-supervision met language-image pre-training, UC Berkeley proposed the multitasking framework SLIP
Pre-training uc .AI Berkeley Clip arxiv
02-09
ICLR 2022 Paper List Published with Up to 32% Acceptance Rate
iclr Deep learning arxiv
01-29
After breaking through 2 million papers, arXiv gradually lost popularity?
arxiv Preprints Luo Huiqian
01-25
Stronger than MAE, FAIR's new method, MaskFeat, refreshes multiple SOTa with HOG
mask below Pre-training Natural language processing arxiv
01-06
AI generates high-number questions, and it is difficult to reach new heights: MIT proposes the first algorithm model that can be asked, done, and scored
artificial intelligence with The heart of the machine Neural networks arxiv
01-04
Github Inventory! The 38 most amazing AI papers of 2021
artificial intelligence https pdf machine learning arxiv almost As swintransformer vision New Wisdom
12-30
Have you read all 40 AI papers you can't miss in 2021?
.AI https code arxiv
12-27
A single ViT model performs multimodal multitasking, and Google implements multiple SOTa with a collaborative training strategy
Google below arxiv
12-21