Come check out our latest ICLR 2021 paper! Congratulations to the leading authors Mihir Prabhudesai, Shamit, Darshan Patil !! Our model can recognize new objects and can answer complex visual questions from a few labeled data.

How well do today's AI models understand the physical structure and dynamics of visual scenes? Introducing #Physion: a dataset/benchmark featuring objects that roll, slide, fall, fold, collide, connect, contain, and more! preprint: arxiv.org/abs/2106.08261

Announcing our NeurIPS Conference workshop titled "Physical Reasoning and Inductive Biases for the Real World". What forms of inductive bias (if any) do we need for physical reasoning in the real world? Submissions due 11 October physical-reasoning.github.io

Advertising our NeurIPS workshop on the topic of physical reasoning and learning! We are trying to bring together people from machine learning, physical sciences, robotics, and psychology to provide new thoughts on this! We welcome you to submit your recent work to our workshop!

Want to know how far are current machines from humans in physical event predictions? How can we go further? Come to our poster at #NeurIPS2021 tomorrow 11:30am - 1pm EST!

Today, raia hadsell, Kyunghyun Cho and I are happy to announce the creation of a new journal: Transaction on Machine Learning Research (TMLR) Learn more in our post: medium.com/@hugo_larochel…

Happening tomorrow! #NeurIPS 2021 workshop on physical reasoning and inductive biases for the real world! Join us at 11 am EST (8 am PST, 4 pm GMT/UTC) For schedule and accepted papers, see physical-reasoning.github.io Fantastic speakers and panelists from industry/academia

Attending #NeurIPS2021 and interested in how to design systems that learn physics for applications like robotics, science, and industry? Then join our workshop tomorrow: Physical Reasoning and Inductive Biases for the Real World!

Much looking forward to this #NeurIPS2021 workshop, Physical Reasoning and Inductive Biases for the Real World, tomorrow at 11am EST. For my part, I'll be talking about this:

I’ll be speaking at this #NeurIPS2021 workshop later today. It’s a great topic and there are many interesting speakers and contributions. A natural follow up to yesterday’s #ML4PS2021

Please join us! It's a wonderfully diverse set of speakers, so I'm sure you'll find something of interest to you

GPT-3 is amazing at complex tasks like creative writing and summarizing. But it's surprisingly bad at reversing words. 🤔 The reason is that GPT-3 doesn't see the world the way we humans do. 👀 If you teach it to reason, it can get around its limitations to get really good. 💡

Nice repeat of piece in the NYT about the importance of early play featuring my favorite studies by Stahl & Feigenson and Carlson et als "batman effect", and our work with Daphna Buchsbaum showing links between pretending and counterfactual thinking. nytimes.com/2020/07/21/par…

After 2 years of work by 442 contributors across 132 institutions, I am thrilled to announce that the github.com/google/BIG-ben… paper is now live: arxiv.org/abs/2206.04615. BIG-bench consists of 204 diverse tasks to measure and extrapolate the capabilities of large language models.

Posted my fifth annual review of my dated predictions from Jan 1, 2018, about self driving cars, AI/ML & Robotics, and human spaceflight. They seemed pessimistic back then, but they are standing the test of time. BONUS five new curmudgeonly predictions. rodneybrooks.com/predictions-sc…

New paper on the topic of solving puzzle boxes and opening visually ambiguous articulated objects, e.g. doors and cabinets, with 3D vision and interaction! The model can solve a puzzle box from < 100 interactions!

Come join us in our CoRL workshop on generalists agents! How can we build robot generalists using learning from video, scaling up simulators, real robot exploration, generative AI and more? #FirstTweet