We have a newly minted PhD student from our group at Princeton University! Congratulations to Dr. Meghan Booker, and good luck on your journey as a researcher at Johns Hopkins APL! 🎓🎓🎓

Celebrating our time with Anushri Dixit , who has spent the year as a postdoc in our lab. She is heading off to UCLA as an Asst. Professor starting July! If you are interested in robot control or related topics, keep an eye out for the amazing work from her new group!

I received tenure at #Princeton! It’s been an incredible journey, and I am immensely grateful to my group, collaborators, mentors, and family. Looking forward to continuing our work on making 🤖 more trustworthy! I will be on sabbatical 2024-25, spending time at Google DeepMind.

Interested in learning more about cross-embodiment learning for robots? Check out Jeannette Bohg's (Jeannette Bohg) fantastic talk here at our Princeton University #Robotics Seminar this Spring! Enabling Cross-Embodiment Learning in Robotics youtube.com/watch?v=tVoQei…

Curious about model-based techniques for controlling humanoid robots? Check out Ye Zhao's great talk in our #Princeton Robotics Seminar earlier this Spring! Unlocking Agility, Safety, and Resilience for Legged Navigation youtu.be/T1kuNXn7-Uw?si…

All recordings from our Princeton University Robotics Seminar are now online! youtube.com/@RoboticsPrinc… We had a wonderful set of speakers spanning diffusion policies, self-supervised learning, human-robot interaction, safe model-based control, and much else. Stay tuned for next year!

I will be spending my sabbatical in the Fall at Google DeepMind in NYC. If you are around and would like to chat or collaborate on robotics, uncertainty quantification, planning, control, or related topics, please ping me!

Engineers at Princeton University and Google, including Anirudha Majumdar, have developed a new way to teach robots to know when they don’t know and ask for clarification from a human. ⬇️ engineering.princeton.edu/news/2024/05/1…

I decided to try my hand at blogging during my sabbatical. First up: "I am your PhD advisor: but what does that actually mean?" On the importance of setting clear expectations in the advisor-advisee relationship, and my version of these in written form. open.substack.com/pub/animajumda…

A typical day of drone research (and the importance of nets).

Six years ago I wrote a post on "Lessons from My First Two Years of AI Research." I got a lot of nice feedback and messages from people over the years saying that it helped. Now that I've finished another career chapter, I've added a few more lessons: tomsilver.github.io/blog/2024/less…

"How many papers do I need to publish during my PhD?": a simple question with a not-so-simple answer. Some thoughts here on why we should think about this question through the lens of a constrained optimization problem in trying to avoid Goodhardt's law: open.substack.com/pub/animajumda…

Anirudha Majumdar - Robots that know when they don’t know: formal assurances for robot learning #conformalprediction

My talk on foundation model-powered robots that know when they don't know at the Princeton University #OLC workshop: youtu.be/23mepB-kkyg?fe… I'm very excited about this work on uncertainty quantification for perception and language.

Informally audit this course from anywhere in the world, for free. Anirudha Majumdar has put his entire Intro to #Robotics course on YouTube.

I’ve always loved my undergrad alma mater Princeton University . But sharing my book with the freshman class of 2028 and President Eisgruber tonight reminded me how much I LOOOVE Princeton, the place where my search for my intellectual North Star of AI began three decades ago!🤩🐯

Diffusion Policy Policy Optimization paper page: huggingface.co/papers/2409.00… We introduce Diffusion Policy Policy Optimization, DPPO, an algorithmic framework including best practices for fine-tuning diffusion-based policies (e.g. Diffusion Policy) in continuous control and robot

How can we go beyond imitation learning with diffusion policies? Check out our new work led by Allen Z. Ren on fine-tuning diffusion policies with policy gradient! Exciting collaboration w/ Max Simchowitz, Toyota Research Institute (TRI), and Pulkit Agrawal 's group. 👇

Accepted to #CoRL2024: our work on calibrated uncertainty quantification for perception systems for end-to-end statistical assurances for navigation! Perceive with Confidence: arxiv.org/abs/2403.08185 Work led by Anushri Dixit during her postdoc at Princeton University!

I feel a huge sense of whiplash in robotics. On one hand, SOTA "generalist" policies don't generalize to even small shifts in initial conditions, distractors, etc. But there are 100s of millions pouring into startups w/ the same technology. Genuine question: what am I missing?