Nataniel Ruiz
@natanielruizg
Followers
8K
Following
10K
Media
355
Statuses
8K
senior research scientist @Google | author of DreamBooth | personalization of generative models (HyperDreamBooth, StyleDrop, ZipLora, RealFill, Unbounded)
Boston, MA
Joined January 2011
language2language.
We're all excited to see our advanced Gemini Deep Think model achieve a gold-medal level performance in the recent IMO, solving 5 of the 6 problems perfectly (35 of 42 points)! 🥇. Unlike our entry last year, which first converted the problems to a formal proof language (Lean),.
0
1
6
it's over guys Gemini is shipping the IMO gold medal model. 10/10 excited.
Our IMO gold model is not just an "experimental reasoning" model. It is way more general purpose than anyone would have expected. This general deep think model is going to be shipped so stay tuned! 🔥.
1
3
103
RT @OriolVinyalsML: Drastic progress on maths with Gemini 2.5! As a math undergrad, I am impressed 🤯. 🥈 -> 🥇 ✅.Formal -> Informal ✅.Special….
deepmind.google
Our advanced model officially achieved a gold-medal level performance on problems from the International Mathematical Olympiad (IMO), the world’s most prestigious competition for young...
0
72
0
huge congrats for this amazing achievement.
An advanced version of Gemini with Deep Think has officially achieved gold medal-level performance at the International Mathematical Olympiad. 🥇. It solved 5️⃣ out of 6️⃣ exceptionally difficult problems, involving algebra, combinatorics, geometry and number theory. Here’s how 🧵
0
1
10
RT @ICCVConference: #ICCV2025 is deeply committed to promoting diversity, equity, and inclusion within our community. As part of this comm….
0
11
0
RT @AbermanKfir: 🎉 Today at @DecartAI we're announcing Mirage. The first generative video model that runs in real-time, to infinity. Mirag….
0
41
0
sometimes i feel like the top labs should be building the new generation of multimodal models with customization/personalization as a base feature. the future is probably one of models that behave differently in deep ways for each user. cracking this seems important.
1
0
9
when llms and multimodal models are able to easily learn the preferences of creators and users in an efficient way there will probably be a revolution on the breadth and depth of tasks than can be tackled by them.
5
1
12
RT @QuentinAnthon15: I was one of the 16 devs in this study. I wanted to speak on my opinions about the causes and mitigation strategies fo….
0
464
0
claude is doing the "it isn't just X - it's Y" now 😢.i don't recall it doing this before.
0
0
1
deep personalization of llms and image/video generative models is going to be a key part of the field (and industry).
2
0
32
RT @swyx: @main_horse metr tried to dunk on ai but accidentally found a great way to interview ai engineers.
0
1
0
RT @ICCVConference: The deadline for camera-ready paper and copyright submission for #ICCV2025 is Friday, August 1st, 2025 @ 11:59 Pacific….
0
7
0
Excited to give a keynote at this workshop.
The first Personalization in GenAI Workshop is coming to @ICCVConference in Hawaii!. World-class speakers & panelists .Cutting-edge research showcase.Network with industry leaders . Don't miss your chance to be part of this groundbreaking event!. ⌛️Deadline: July 07 - Submit.
0
1
15
cool data point.
working with $GOOG JAX Pallas, my immediate reaction is that this is Python CUDA. got $AMD MI3xx chips to work quite easily, dense matmuls are fast, but significant effort is needed to tune whole graph perf. IMO, the $NVDA CUDA moat is disappearing, but the total $NVDA ecosystem.
0
0
2
my talk on generative video games at the CVEU workshop at CVPR, timestamped (thank you Victor).
@cveu_workshop @natanielruizg Did you miss Nataniel's talk? 👇👇👇.
1
0
21
RT @YutongBAI1002: What would a World Model look like if we start from a real embodied agent acting in the real world?. It has to have: 1)….
0
123
0
magic insert has been accepted to ICCV 🥳.still believe a lot in two step processes like this one for some applications to achieve really standout results.
With friends at @Google we announce 💜 Magic Insert 💜 - a generative AI method that allows you to drag-and-drop a subject into an image with a vastly different style achieving a style-harmonized and realistic insertion of the subject (Thread 🧵).web:
8
3
64