My first tweet and it's an exciting one! I'm thrilled to announce that I'm joining CMU Robotics Institute as an Assistant Prof. this Fall. If you are interested in the future of learning-enabled robots that can safely work *with* and *for* people, I'm always looking for new collaborators.

Really exciting to see multi-agent safety methods scale to 200-D and actively account for SOTA data-driven human models, like neural network motion predictors. Check out our new work for Conference on Robot Learning !

It was really exciting to see how our conformal controller has the emergent ability to ignore irrelevant uncertainty, only accounting for that which affects decisions. Looking forward to more theoretical and algorithmic work on decision calibration!

Excited announce the 1st Annual Northeastern Systems and Control Workshop (NESCW), which will be held Penn Electrical and Systems Engineering on May 4th, 2024.   If you are a systems and controls researcher in the northeast, you are invited! nescw.seas.upenn.edu IEEE Control Systems Society IFAC_Control L4DC Conference

And registration is now open! Registration link below and available on the workshop website! forms.gle/YGzwfyFqJC5Vet…

🚨🚨Registration deadline of April 1, 2024🚨🚨 Don't forget to register for the Northeast Systems and Control Workshop, happening at Penn on May 4th, 2024! Register here: forms.gle/YGzwfyFqJC5Vet… More info here: nescw.seas.upenn.edu

Thrilled to receive the Google Research Scholar Award! Thanks to Google AI, we will push the frontier of safe and scalable robot behavior alignment from human feedback. 🤖🤝👩 research.google/programs-and-e…

Really excited about this! In IRL, we know that robot learners should match the visual features in expert demos. But, we found that off-the-shelf vision encoders are not enough. We need human feedback to learn the "right" visual features a robot should match.

Honored and excited to be a mentor this year! Stop by from 12:45p-1:30p today at #ICLR2024 to chat about all things related to learning, robotics, AI safety, academia and more.

It took a lot of work over multiple years but I'm proud to be able to finally share my first #ICML2024 paper: a new, fundamental framework for designing efficient inverse RL algorithms that we're in the process of scaling to real robots: gokul.dev/hyper/! 1/

Human trajectory prediction often assumes consistent behavior trends over time. But human behavior is transient: party-goers act differently at night than when going to work. To address this, our ECCV ’24 paper introduces “latent corridors” for rapid deployment scene adaptation.

5 robotics professors (CMU, Stanford, Cornell). no prior coordination. pure joy. We are going to start an ICRA/IROS/RSS etc dress day for all who wish to participate. From left Zeynep Temel Henny Admoni Jeannette Bohg Andrea Bajcsy and me #RSS2024

We’re so proud to have Emily on our team! x.com/adeleswomen/st…