“Make it red.” “No! More red!” “Ughh… slightly less red.” “Perfect!” ♥️ 🎚️Kontinuous Kontext adds slider-based control over edit strength to instruction-based image editing, enabling smooth, continuous transformations!

Excited to introduce DiffuseNNX, a comprehensive JAX/Flax NNX-based library for diffusion and flow matching! It supports multiple diffusion / flow-matching frameworks, Autoencoders, DiT variants, and sampling algorithms. Repo: github.com/willisma/diffu… Delve into details below!

I used to think that semantic encoders primarily captured high-level, abstract representations and discarded fine-grained visual details, but I was wrong. we employ pretrained representation encoders (such as DINO, SigLIP, and MAE, all based on standardized ViTs) combined with

as always, we’re releasing everything: the paper, the model, and the PyTorch code. this project has been led by three amazing students: Boyang Boyang Zheng (1st year phd), Willis Willis (Nanye) Ma (2nd year phd), and Peter Peter Tong (3rd year phd). we’ve been working on this for

✨ I’ll be presenting our work on depth-aware image editing at #ICCV2025 in Hawaii 🌴 next week! 📅 Oct 22 | 📍 Exhibit Hall I | 🧩 Poster #82 🌍 Project: rishubhpar.github.io/DAEdit/ 🤝 Working on image generation or editing? I’d love to chat at ICCV! Vision and AI Lab, IISc

Full video of the ICCV '25 presentation

Respect. 🫡🚀 Sundar Pichai

Can we scale thinking without scaling context budget!? In our latest work, we propose a test-time scaling strategy which combines parallel drafts in a sequential self-improvement loop to further boost reasoning capabilities of frontier LLMs such as O3 and Gemini-2.5-flash

Our bigger group at Google DeepMind is hiring interns for next summer! If you are interested in working with us, apply through the link below and also email us. step 1: google.com/about/careers/… (US-based) step 2: send an email to [email protected]

ok, I know I’m the biggest stan — but amongst many lessons I have learned from Pat Grady is that cynicism/selfishness is for losers, and openheartedness is for winners. believing in the talented people around you, assuming good intent, playing for your team - it is often enough!

Veo 3.1 now has a Camera Adjustment Feature, allowing you to change the angle and movement of a previously generated video. Taking it out for a test spin, here's our "Test" video, in the thread we'll check out how the feature does!

We're looking for Summer Interns to join the Post-Training Team at @NVIDIA! DM me with your updated resume and three concise bullets detailing your most relevant experience — e.g. publications, repos, blogs, etc. RT please to help us find top talent.

One of the best talks I've ever had for World Model. Definitely worth watching!

I'm looking for a student researcher to work with me at Google DeepMind in London, preferably starting early next year -- topics will be around novel video model architectures / learning from a single video stream / representation learning .

Introducing Cambrian-S it’s a position, a dataset, a benchmark, and a model but above all, it represents our first steps toward exploring spatial supersensing in video. 🧶

MLLMs are great at understanding videos, but struggle with spatial reasoning—like estimating distances or tracking objects across time. the bottleneck? getting precise 3D spatial annotations on real videos is expensive and error-prone. introducing SIMS-V 🤖 [1/n]

Wow! Sometimes you're lucky to wake up and see something fundamentally different to address the challenge of modeling how humans can stream continuous frames of visual information and store infinite context! Exciting times researching video and spatial supersensing! ✨🫡

I wish that everyone gets to experience running the NYC Marathon once in their lives, every human deserves to experience something like it, and it made all the training, discipline, and suffering along the way this year worth it.

🎥 Video Generation Enables Zero-Shot Robotic Manipulation 🤖 Introducing PhysWorld, a framework that bridges video generation and robot learning through (generated) real-to-sim world modeling. 🌐 Project: pointscoder.github.io/PhysWorld_Web/ 📄 Paper: arxiv.org/abs/2511.07416 💻 Code:

carmel by-the-sea is one of the best places on earth debates welcome <3