Zobrazeno 1 - 4
of 4
pro vyhledávání: '"Kumar, Tanishq"'
The existence of "lottery tickets" arXiv:1803.03635 at or near initialization raises the tantalizing question of whether large models are necessary in deep learning, or whether sparse networks can be quickly identified and trained without ever traini
Externí odkaz:
http://arxiv.org/abs/2402.01089
We propose that the grokking phenomenon, where the train loss of a neural network decreases much earlier than its test loss, can arise due to a neural network transitioning from lazy training dynamics to a rich, feature learning regime. To illustrate
Externí odkaz:
http://arxiv.org/abs/2310.06110
Autor:
Zhang, Mengmi, Dellaferrera, Giorgia, Sikarwar, Ankur, Chen, Caishun, Armendariz, Marcelo, Mudrik, Noga, Agrawal, Prachi, Madan, Spandan, Shetty, Mranmay, Barbu, Andrei, Yang, Haochen, Kumar, Tanishq, Han, Shui'Er, Singh, Aman Raj, Sadwani, Meghna, Dellaferrera, Stella, Pizzochero, Michele, Tang, Brandon, Ong, Yew Soon, Pfister, Hanspeter, Kreiman, Gabriel
As AI algorithms increasingly participate in daily activities, it becomes critical to ascertain whether the agents we interact with are human or not. To address this question, we turn to the Turing test and systematically benchmark current AIs in the
Externí odkaz:
http://arxiv.org/abs/2211.13087
Autor:
Zhang, Mengmi, Dellaferrera, Giorgia, Sikarwar, Ankur, Armendariz, Marcelo, Mudrik, Noga, Agrawal, Prachi, Madan, Spandan, Barbu, Andrei, Yang, Haochen, Kumar, Tanishq, Sadwani, Meghna, Dellaferrera, Stella, Pizzochero, Michele, Pfister, Hanspeter, Kreiman, Gabriel
As AI algorithms increasingly participate in daily activities that used to be the sole province of humans, we are inevitably called upon to consider how much machines are really like us. To address this question, we turn to the Turing test and system
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::b3c58fc0f8a5b84cb32509981c174b55