1/ Commonsense reasoning needs multimodal knowledge, yet current LLMs focus mostly on text, limiting their integration of crucial visual information. We introduce vLMIG, a method that enhances LLMs' visual commonsense by integrating images into the decision-making process

I am excited to share my first work: "Dataset Size Recovery from LoRA Weights". Ever wondered if you could find out how many samples was a model trained on using just its weights? Well now you can! Project: vision.huji.ac.il/dsire/ 👇

ObjectDrop is accepted to #ECCV2024! 🥳 In this work from Google AI we tackle photorealistic object removal and insertion. Congrats to the team: Matan Cohen Shlomi Fruchter Yael Pritch  Alex Rav-Acha Yedid Hoshen Checkout our project page: objectdrop.github.io

One last poster I forgot to post. «Recovering the Pre-Fine-Tuning Weights of Generative Models» by Eliahu Horwitz #ICML2024 Clear explanation, 8 line code, visually appealing, and QR code for additional information.

Google presents Diffusion Models Are Real-Time Game Engines discuss: huggingface.co/papers/2408.14… We present GameNGen, the first game engine powered entirely by a neural model that enables real-time interaction with a complex environment over long trajectories at high quality.

I'm sharing something unique we've been making at Google (w/ UNC). We are releasing our work on a new class of interactive experiences that we call generative infinite games, essentially video games where the game mechanics and graphics are fully subsumed by generative models 🧵

🚨 Excited to share ObjectMate our latest from Google AI for zero-shot subject-driven generation & insertion 🚨 🔗 project page: object-mate.com 📄Arxiv: arxiv.org/abs/2412.08645

Happy to introduce Veo 2, our latest video generation model, rolling out to trusted testers today. It's not perfect, but it's a significant improvement over current state-of-the-art models, as our benchmarks demonstrate.

🚨 New paper alert! 🚨 Millions of neural networks now populate public repositories like Hugging Face 🤗, but most lack documentation. So, we decided to build an Atlas 🗺️ Project: horwitz.ai/model-atlas Demo: huggingface.co/spaces/Eliahu/… 🧵👇🏻 Here's what we found:

Our work maps hereditary relationships between models. We find that weights🏋️ are sufficient for decoding the Origin of Models🌳 Presenting today at ICLR 2026, 15:00–17:30, Hall 3, poster #360. Come by to see our method, visualizations, and interactive demo atlas👀🚀 #ICLR2025

Google presents LightLab Controlling Light Sources in Images with Diffusion Models

A quick look on what we were working on in the past months 🎉 more exciting things to come 🙌🏻

Prompt Theory (Made with Veo 3) What if AI-generated characters refused to believe they were AI-generated?

Used Veo 3 to imagine if Jurassic Park was real. This entire thing is AI! Things are gonna get weird #Veo3 #ai #JurassicPark #JurassicWorld

Andrej Karpathy Thanks for the inspiring talk (as always!). I'm the author of the Model Atlas. I'm delighted you liked our work, seeing the figure in your slides felt like an "achievement unlocked"🙌Would really appreciate a link to our work in your slides/tweet arxiv.org/abs/2503.10633

We are releasing a paper I'm very excited about. We know test-time scaling is a path to greatly improved results, and achieves reasoning in the case of LLMs. We present a new and promising way to amortize it into training using HyperNetworks for image generation models.

Image generation with Gemini just got a bananas upgrade and is the new state-of-the-art image generation and editing model. 🤯 From photorealistic masterpieces to mind-bending fantasy worlds, you can now natively produce, edit and refine visuals with new levels of reasoning,

Excited to share this has now been accepted at #NeurIPS2025 as a position paper (<6% acceptance)!🎉 We advocate for systematically studying entire model populations via weight-space learning, and argue that this requires charting them in a Model Atlas. NeurIPS Conference #NeurIPS 🧵👇