Navigate Select ESC Close

DeepSeek's Deleted Vision Paper Is Nuts...

2026-07-08 Science & Technology
3.2k
317
22
bycloud
bycloud
229.0k subscribers

Unlock all features

FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.

Description

Need high quality cloud GPUs? Check out Verda now, and use the code BYCLOUD-50 to get $50 of compute credits for just $5! https://verda.com/?utm_source=bycloud&utm_medium=referral&utm_campaign=sponsorship&utm_content=deepseekv4 How to redeem the discount: https://docs.verda.com/resources/obtaining-free-credits/how-to-redeem-credits/#redeem-discount-coupons In this video, we are diving into the DELETED paper from DeepSeek called Visual Primitives, which is a vision model that grounds vision through pointers and boxes. my latest project: Intuitive AI Academy We just wrote a new piece on Optimization!! https://intuitiveai.academy/ limited time code "LOCKIN" for 35% off yearly plan My Newsletter (weekly top research papers) https://mail.bycloud.ai/ My Patreon https://www.patreon.com/c/bycloud SOURCES: Thinking With Visual Primitives [Paper] https://www.alphaxiv.org/abs/visual-primitives Try out my new fav place to learn how to code https://scrimba.com/?via=bycloudAI This video is supported by the kind Patrons & YouTube Members: 🙏Spam Maj, Alex, Chris LeDoux, DX Research Group, Poof N' Inu, Deagan, Robert Zawiasa, Ryszard Warzocha, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa, Toru Mon, Lame Plane, Matej Macak, Len Mo, saylikhapekar, ZyanSheep, THEVIERAOS, Ricardo Raphael Corona-Moreno, superchordate [Discord] https://discord.gg/NhJZGtH [Twitter] https://twitter.com/bycloudai [Patreon] https://www.patreon.com/bycloud [Business Inquiries] [email protected] [Other Inquiries] [email protected] [Profile & Banner Art] https://twitter.com/pygm7 [Video Editor] Abhay Manim Animations created with Manimate https://www.manimate.ai/ [Ko-fi] https://ko-fi.com/bycloudai [Bitcoin (BTC)] 3JFMJQVGXNA2HJE5V9qCwLiqy6wHY9Vhdx [Ethereum (ETH)] 0x3d784F55E0bE5f35c1566B2E014598C0f354f190 [Litecoin (LTC)] MGHnqALjyU2W6NuJSSW9fTWV4dcHfwHZd7 [Bitcoin Cash (BCH)] 1LkyGfzHxnSfqMF8tN7ZGDwUTyBB6vcii9 [Solana (SOL)] 6XyMCEdVhtxJQRjMKgUJaySL8cGoBPzzA2NPDMPfVkKN

Top Comments (10)

@Lyoishi 2026-07-08

Extremely efficient Vision models sound like exactly what self driving wants

204 6 replies
@CarsMeetsBikes 2026-07-08

so vision is basically a graph problem? something something always has been...

100 6 replies
@dragonmares59110 2026-07-08

I truly wish some groups would focus on little models, for embedded, for specialized task or subject or for low energy application. We don't all need giant toolbox, sometime a simple thing can do the job too.

82 5 replies
@TheBayHarbourPub 2026-07-08

Vision models are so interesting😮

52 3 replies
@GargoilXD 2026-07-09

Closed Source Models: yeah we're gonna copy that, thanks.

18 2 replies
@Kknewkles 2026-07-08

Heh, I prudently downloaded this paper the last time you mentioned it :DDD

9
@hyperefract 2026-07-09

YESSSS ACTUAL IMAGE POSITIONS

7
@bycloudAI 2026-07-08

Need high quality cloud GPUs? Check out Verda now, and use the code BYCLOUD-50 to get $50 of compute credits for just $5! https://verda.com/?utm_source=bycloud&utm_medium=referral&utm_campaign=sponsorship&utm_content=deepseekv4 How to redeem the discount: https://docs.verda.com/resources/obtaining-free-credits/how-to-redeem-credits/#redeem-discount-coupons

5 2 replies
@steve_jabz 2026-07-08

ARC Challenge: Left the chat.

5
@midsingularity 2026-07-08

Bro seeing this thumbnail as an AI enthusiast and a Perc Angle mark just blew my mind.

1

Unlock the Data Inside
Turn Videos into Knowledge

  • Get FREE 10/day: transcripts, summaries, chats
  • Chat with videos, export text & PDF
  • $1 free API credit for RAG, chatbots & research

Free forever plan • All features unlocked

App screenshot