🔥Telegram İfşa

Excited to see our paper "Tokenformer: Rethinking transformer scaling with tokenized model parameters" accepted as a spotlight at #ICLR2025 ! Hope our idea of ​​tokenizing everything can inspire the future of AI. Paper: arxiv.org/abs/2410.23168 Code: github.com/Haiyang-W/Toke…

Excited to share what we have been up to in image text embedding models. SigLIP 2 is the most powerful encoder for most open vocabulary computer vision and MMLLM tasks. Checkpoints are open sourced and we look forward to what the community achieves with these.

📢2⃣ Yesterday we released SigLIP 2! TL;DR: Improved high-level semantics, localization, dense features, and multilingual capabilities via drop-in replacement for v1. Bonus: Variants supporting native aspect and variable sequence length. A thread with interesting resources👇

Delighted to share that ACED has been accepted at CVPR2025! Check out our work to know how to distill the strongest smol size image-text contrastive models.

Fully supportive of this. Machine Learning/ Computer Vision review process is broken with irresponsible reviewers. Glad to see there is some accountability.

The majority of features in this layer of Siglip-2 are multimodal. I'd expected some multimodality but was surprised that two-thirds of the neurons I tested bind together their visual and linguistic features. This neuron fires for images of mustaches and for the word "mustache"

Check out the strongest open-source dense prediction models from our colleagues!

Google's global PhD Fellowship program will open for applications this week! (on Apr 10th) This supports PhD students in computer science and related fields, also connecting to a Google mentor. Learn more and apply at: goo.gle/phdfellowship (deadline: May 15th, 2025)

We are presenting JetFormer at ICLR this morning, poster #190. Stop by if you’re interested in unified multimodal architectures!

📢 We just released the code for JetFormer at github.com/google-researc… Enjoy!

At #GoogleIO, we shared how decades of AI research have now become reality.  From a total reimagining of Search to Agent Mode, Veo 3 and more, Gemini season will be the most exciting era of AI yet.  Some highlights 🧵

timm's got a new vision transformer (NaFlexVit), and it's flexible! I've been plugging away at this for a bit, integrating ideas from FlexiViT, NaViT, and NaFlex and finally ready to merge for initial exploration. The model supports: * variable aspect/size images of NaFlex (see

Stop by this amazing work from Vishaal and the team today at CVPR

Active Data Curation Effectively Distills Large-Scale Multimodal Models - compute per sample loss with large batch - only backprop (probabilistically) through samples with high loss intuition: these are the samples where there is “something to learn” - if both teacher and

A big congratulations to the whole Gemini team on pushing this amazing family of models out 😄 Our tech report is out now: storage.googleapis.com/deepmind-media… Feels a bit unreal to share the contributors list with all the amazing colleagues

Oguzhan is an amazing mentor to work with, apply if you are on the internship market

🎺Meet VIST3A — Text-to-3D by Stitching a Multi-view Reconstruction Network to a Video Generator. ➡️ Paper: arxiv.org/abs/2510.13454 ➡️ Website: gohyojun15.github.io/VIST3A/ Collaboration between ETH & Google with Hyojun Go, Dominik Narnhofer, Goutam Bhat, Federico Tombari, and Konrad Schindler.

🚀 Excited to share our new work RefAM: Attention Magnets for Zero-Shot Referral Segmentation, a training-free approach that turns diffusion model attentions into segmentations. By Anna Kukleva, me, Alessio Tonioni, Ferjad Naeem, Federico Tombari, Jan Eric Lenssen, Bernt Schiele

First Gemini release with a piece of my work inside 😄 and all the countless other amazing people 🚀🚀🚀

(3/5)2016-2021 was a wonderful period for AI research precisely because the leading labs at the time – FAIR, Deep Mind, Google, Open AI – were all publishing freely and building off each other’s results. If the transformer paper in 2017 had been held as a secret inside Google