Have you ever wondered what makes two models different? We all know the ViT-Large performs better than the Resnet-50, but what visual concepts drive this difference? Our new ICLR 2025 paper addresses this question! nkondapa.github.io/rsvc-page/

So cool to see CoT for images from #OpenAI's new release today! Congrats Bowen Cheng and team! We tried it on a few examples from our new and very challenging 3D reasoning benchmark, Omni3D-Bench (at #CVPR2025) and found some really interesting reasoning outputs 🤩👇

A must-read for anyone working on 3D perception & LLMs for 3D!! 👀 Congrats on the release! Aadarsh Sahoo

I’ll be presenting VADAR at #CVPR2025 on Saturday, June 14 from 17:00-19:00 at Poster 347! Come stop by & talk about 3D visual reasoning for LLMs and visual programming!! 👋👋

Would love to meet more people #CVPR2026!! If you’d like to chat over coffee please reach out! #CVPR2025

In many vision tasks, perceptual reasoning does not come naturally. Experts still have to deeply study an image, deduce relevant concepts, and reason about them in natural language (inaturalist.org/observations...). Our goal is to automate this process – with no human oversight.

I had to cancel my trip to #CVPR2025 because I caught the stupid flu 🙃 So sad to miss everyone! But if you are in Nashville go check out Damiano's project tomorrow on improving 3D spatial reasoning from a single image with the power of LLMs 🌟 #CVPR2025 Damiano Marsili

See you next year in Denver!

Had a great time at #CVPR2025 showcasing VADAR!! Thank you to everyone who stopped by the poster & all the great conversations 🙌🙌 #CVPR2025

Inspired time Laude Institute, huge congrats to Andy Konwinski on the launch! Humbled to be asked to speak on the dos of impact research in front of such 🤩 researchers!! P.S. 👀 the where's waldo moment of my favorite +1 (Yisong Yue) being (I hope 😅) a proud academic dad 🪞📸

🥋 We're excited to share judo: a hackable toolbox for sampling-based MPC (SMPC), data collection, and more, designed to make it easier to experiment with high-performance control. Try it: pip install judo-rai

Super excited about 🥋judo🥋, our new open-source sampling-based MPC toolbox! 🖥️Try in under 30s: pip install judo-rai judo <click link> 🧵thread below👇

🚀 Sharing our #ICCV2025 paper, "Is CLIP ideal? No. Can we fix it? Yes!". We will be at Poster session 5 (10:45AM 10/23), please come find us to chat or reach out online! A thread: [1/6]

I am at ICCV this week giving talks tomorrow (10/20) at: ✴️What's next in Multimodal Foundation Models (326B, 11.10am) on how we can integrate 3D into a next-token prediction framework along with images and text ✴️Closing the Loop Between Vision and Language (313C, 2.45pm)

Perona and Belongie wondered if they could replace search words with images, but they knew that it would not be easy to convince people to tediously and appropriately tag every object and its parts in millions of pictures. caltech.edu/about/news/pie…

1/9 Videos of the real world are chaotic: filled with occlusions, abrupt motions, and non-rigid deformations. How well can current tracking models handle these difficulties? In our NeurIPS paper we answer this question with ITTO, a challenging new benchmark suite for evaluating

Since she's way too shy to post this herself, please join me in congratulating my amazing colleague and friend Katie Bouman for receiving tenure at @caltech! 🥳🎉

Excited to share our paper Representational Difference Explanations (RDX) was accepted to #NeurIPS2025! 🎉RDX is a new method for model diffing designed to isolate 🔍 representational differences. 1/7

3Dfy anything from a single image! Very thrilled to announce SAM 3D. From an input image, select any object you want, 3Dfy it! Blog: ai.meta.com/blog/sam-3d/ Demo: aidemos.meta.com/segment-anythi…