In emergencies, minutes can decide who lives and who dies. Our team is participating in the Triage Challenge, building AI to empower clinicians and prioritize care when resources are thin:
prontotriage.com

Also recently featured by Meta:
ai.meta.com/blog/upenn-d...

(1/5)

🎉 Announcing the winners of the #CoLLAs2025 Early-Career Spotlight Program!

These rising stars are pushing the boundaries of lifelong & continual learning, adaptation, and embodied AI.

👇 A thread introducing their work & talks.

#AI #MachineLearning #ContinualLearning #LifelongLearning

It was lots of fun moderating this theatrical, wrestling-themed debate on CS vs AI degrees at the 2025 CRA Summit. Thanks to my panelists for being such good sports and having great insights! #CRASummit
@vasant-h.bsky.social

Intersectional Fairness in Reinforcement Learning with Large State and Constraint Spaces In traditional reinforcement learning (RL), the learner aims to solve a single objective optimization problem: find the policy that maximizes expected reward. However, in many real-world settings, it ...

Come say hi at our #ICML poster today during Poster Session 1 (W-600)! Joint work with @ericeaton.bsky.social @marcelhussing.bsky.social @optimistsinc.bsky.social @aaroth.bsky.social @mkearnsphilly.bsky.social!

arxiv.org/abs/2502.11828

📢 Only 8 weeks to #CoLLAs2025!

Tackle the real-world challenges of ML: catastrophic forgetting, dist. shift, continual RL, lifelong learning & more.

📍 Aug 11–14 @upenn.edu
🧠 Talks, workshops, posters, tutorials & community

🔗 lifelong-ml.cc/Conferences/...

#ContinualLearning #AI #MachineLearning

MAD-TD: Model-Augmented Data stabilizes High Update Ratio RL Building deep reinforcement learning (RL) agents that find a good policy with few samples has proven notoriously challenging. To achieve sample efficiency, recent work has explored updating neural...

I'll be at @iclr-conf.bsky.social next week! Ping me if you want to chat about RL, life, or the universe! I'll be presenting our MAD-TD paper openreview.net/forum?id=6Rt... at the poster session, and I will also most likely be lurking at the Robot Learning workshop.

Do you want to get the most out of your samples, but increasing the update steps just destabilizes RL training? Our #ICLR2025 spotlight 🎉 paper shows that using the values of unseen actions causes instability in continuous state-action domains and how to combat this problem with learned models!

Reposting a postdoc opportunity in field robotics at Penn, in case anyone missed it the first time around.

Congratulations! Ignore any naysayers; it’s still an honor. In case it helps, I got some guidelines from Michael Littman years ago and now send them to all students who request a recommendation. It helps smooth the process a lot: www.cis.upenn.edu/~eeaton/reco...

This is so sad. It’s also yet another reason why I strongly believe that hardware shouldn’t be completely dependent on cloud services. It’s fine as a value-added on top of local functionality, but loss of the cloud service, or even the internet, should never brick a device.

Dissecting Deep RL with High Update Ratios: Combatting Value Divergence We show that deep reinforcement learning algorithms can retain their ability to learn without resetting network parameters in settings where the number of gradient updates greatly exceeds the number o...

Been seeing a lot of RL architectures use layer norms as "design choice" without elaborating why it's needed. If you are curious take a look at our paper: arxiv.org/abs/2403.05996 TLDR: Layer norms stop your critic from diverging early on and stabilize training. cc @cvoelcker.bsky.social

A starter pack of researchers in AI and ML at Penn, currently woefully incomplete. Please PM to be added or suggest names.

go.bsky.app/5SxWsoZ

New feed on continual learning, lifelong learning, and related ML topics: bsky.app/profile/did:...

I hope you find it useful. Please let me know if you notice it is missing any posts, so that I can adjust the regular expression.

Come join us in Philadelphia in August for CoLLAs 2025!

We're hiring a postdoc in field robotics at the University of Pennsylvania's GRASP lab, which will focus on multi-robot autonomy, control/coordination, and perception for the DARPA Triage Challenge. #robotics

For details and to apply, see www.seas.upenn.edu/~eeaton/open...