Sujoy Paul
@paul_sujoy_
Followers
50
Following
46
Statuses
19
Research Scientist @ Google DeepMind
India
Joined June 2023
Try it out! It’s quite good.
🎥 Our state-of-the-art video generation model Veo 2 is now available in @YouTube Shorts. With the Dream Screen feature, creators can: ✨ Produce new clips that fit seamlessly into their storytelling with a quick text prompt ✨ Use it to make backgrounds for their videos. Here’s how →
0
0
2
RT @adityakusupati: What can be more Matryoshka 🪆 than the bit structure of an integer data type! Upto 10% better results for low precisio…
0
11
0
Check out our latest work on inference time scaling of model size for image generation!
Can we generate high-quality images/videos 3× faster without extra parameters? 🤯 Our latest work “MaGNeTS” rethinks compute efficiency in generative models using decode time model scaling & KV caching!
0
2
6
Stop by our poster! We will be talking about adaptive compute. @gaganjain1582 is also applying for PhD, and I have only great things to say about him. Also stop by #2507 to learn about elastic models by the amazing @adityakusupati and co!
It's todayyy! If you are at NeurIPS, come to our poster (#2502) in the East Exhibit Hall at 4.30 PM. I'll also be at the Google DeepMind booth from 3 to 4 PM. #NeurIPS #NeurIPS2024
0
1
12
RT @JeffDean: 🎊Gemini 2.0 is here! 🎊 An AI model for the agentic era The blog post is chock full of announcements: Gemini 2.0 Flash Proj…
0
139
0
RT @gaganjain1582: Give me your tokens and I'll process them faster :D Here's our #NeurIPS poster. Excited to hear your thoughts and answe…
0
11
0
RT @partha_p_t: Application are now open for the Predoc 2025 program at Google DeepMind India. If you are pre-PhD (i.e., current undergrad…
0
65
0
RT @JeffDean: This team is doing amazing work! I visited them in February and had some of the most enjoyable research reviews I've ever do…
0
9
0
RT @gaganjain1582: MoNE makes it to #NeurIPS2024 @NeurIPSConf! Thanks everyone for the incredible love and support ❤️ Super happy! @paul_s…
0
6
0
RT @adityakusupati: Extremely proud of MoNE -- that enables conditional compute on visual understanding (image/video) tasks which have a lo…
0
3
0
RT @TheAITimeline: 🚨This week’s top AI/ML research paper (Meta vs DeepMind!): - SAM 2 (Meta FAIR) - Gemma-2 (DeepMind) - MoMa (Meta FAIR) -…
0
15
0
RT @Ar_Douillard: Mixture of Nested Experts: Route tokens to encoder transformer blocks of different sizes. Harder…
0
28
0
RT @gaganjain1582: Your models about to get a MoNE-y makeover! 💸 We're thrilled to unveil MoNE (Mixture of Nested Experts), a game-changer…
0
14
0
RT @_akhaliq: Google presents Mixture of Nested Experts Adaptive Processing of Visual Tokens The visual medium (images and videos) natura…
0
77
0
RT @gaganjain1582: 👀A sneak-peek at our recent work accepted at #ECCV2024 @eccvconf ! 📜: Want a scalable and robu…
0
23
0