Amir Globerson (@amirgloberson) 's Twitter Profile
Amir Globerson

@amirgloberson

ID: 249189776

calendar_today08-02-2011 15:22:47

19 Tweet

131 Followers

228 Following

Amir Bar (@_amirbar) 's Twitter Profile Photo

Excited to share DETREg, our new work on unsupervised pretraining for object detection with transformers. Compared to previous works, the key idea in DETReg is attempting to learn object detection in the unsupervised pre-training stage. Berkeley AI Research , Tel Aviv University, NVIDIA AI

Excited to share DETREg, our new work on unsupervised pretraining for object detection with transformers. Compared to previous works, the key idea in DETReg is attempting to learn object detection in the unsupervised pre-training stage. <a href="/berkeley_ai/">Berkeley AI Research</a> , <a href="/TelAvivUni/">Tel Aviv University</a>, <a href="/NVIDIAAI/">NVIDIA AI</a>
Roei Herzig (@roeiherzig) 's Twitter Profile Photo

The work 🖥️ will be presented today from 2:30 pm to 5:00 pm in Hall B2-C as part of the Poster 1.2 session #CVPR2025. Come to chat with us or even just stop by and say hello!👋 #CVPR2022 #CVPR22

Ori Ram (@ori__ram) 's Twitter Profile Photo

#NLProc Dense retrievers are great! But how do they encode their inputs? "What Are You Token About? Dense Retrieval as Distributions Over the Vocabulary" We show that projecting dense retrieval representations to the vocabulary space helps understand and improve them! 🚀 A 🧵

#NLProc
Dense retrievers are great!
But how do they encode their inputs?

"What Are You Token About? Dense Retrieval as Distributions Over the Vocabulary"

We show that projecting dense retrieval representations to the vocabulary space helps understand and improve them! 🚀

A 🧵
Yuval Kirstain (@ykirstain) 's Twitter Profile Photo

I am super excited to introduce the Pick a Pic community effort in text-to-image generation! pickapic.io Pick a Pic is a completely *open-source* and *collaborative* project to align text-to-image generation with what humans really *want* to see. 🧵 1/3

I am super excited to introduce the Pick a Pic community effort in text-to-image generation!
pickapic.io 
Pick a Pic is a completely *open-source* and *collaborative* project to align text-to-image generation with what humans really *want* to see. 
🧵 1/3
Amir Globerson (@amirgloberson) 's Twitter Profile Photo

#NeurIPS2023 needs you! If you received an invitation to review for NeurIPS 2023, please accept it. If you have students, ask them to accept their invitations. Thanks!

NeurIPS Conference (@neuripsconf) 's Twitter Profile Photo

Our past conferences wouldn't have been possible without our many reviewers. If you have at least 2 papers in top peer-reviewed confs or journals, with at least one in an ML venue (eg NeurIPS ICML ICLR), we’d be v grateful if you reviewed for #NeurIPS2023 forms.gle/3bAjsYtD5FEpxJ…

Mor Geva (@megamor2) 's Twitter Profile Photo

LMs capture many factual associations, but how do they recall them internally during inference? In a new preprint, we find that LMs build attribute-rich subject representations, from which attention heads extract the predicted attribute. Jasmijn Bastings Katja Filippova Amir Globerson 🧵

LMs capture many factual associations, but how do they recall them internally during inference?
In a new preprint, we find that LMs build attribute-rich subject representations, from which attention heads extract the predicted attribute.
<a href="/jasmijnbastings/">Jasmijn Bastings</a> <a href="/fajtak/">Katja Filippova</a> <a href="/amirgloberson/">Amir Globerson</a> 🧵