Speaking in 30min about agents and safety in Computer use agents at West Meeting room 211-214 #ICML2025

Come listen to the second talk in 15min about the role of data in building trustworthy LLMs and memorization/creativity in LLMs. #ICML2025 Where? Ballroom West A

Intelligence isn't a collection of skills. It's the efficiency with which you acquire and deploy new skills. It's an efficiency ratio. And that's why benchmark scores can be very misleading about the actual intelligence of AI systems.

life update: I'll be starting my PhD in CS at Stanford this September! I'm very excited to continue my research on reasoning of language models and to make new friends in the Bay Area! I'm deeply grateful to everyone who supported me and made this milestone possible

Our #ACL2025NLP workshop REALM on LLM agents is happening July 31 in Vienna 🎶🎼 🗓️ Schedule & accepted papers are live! realm-workshop.github.io 🚀Join us for a day of invited talks, paper presentations and a panel discussion with an amazing line-up!

No way🤣🤣 I thought this is a joke but it's VERY true🥹 And it's neurips btw😱🤣

Come join us tomorrow at the 1st LLM agents workshop in ACL (REALM); amazing talks and oral presentations are ahead, with an incredible lineup of speakers🔥🔥 #ACL2025 #ACL2025NLP 🇦🇹🇦🇹 Where? Room 1.61-62 When? tomorrow Thursday (starting 8:45am) realm-workshop.github.io

What's the path to scalable and safe web agents? Is web agents the new semantic parsing? I will be giving a talk at the ACL REALM workshop today at 9:30 am. Come check out if you are interested in the history and contemporary work in this area. Lot of other exciting speakers.

The workshop is continuing ... Roberta Raileanu gave an incredible talk about different ingredients on how to build AI scientist agent and explained how current systems lag far behind human performance in terms of creativity and discovery #ACL2025

REALM is continuing in the afternoon with Tao Yu, Yu Su and Daniel Fried, taking about: - Open Foundations for Computer-Use Agents - Augmenting Human Cognition with AI Agents that Use Computers - Inducing Functions to Improve LLM Agents #ACL2025 #acl2025nlp

The crowd was so huge that the workshop had to be broadcast in the hallway outside 🥹 I hope next time we get a bigger room (and more chairs outside too) so people don't have to stand for so long whether inside or out. Chris Manning gave an amazing thoughtful talk about how can

Huge thanks to the organizers for putting together a fantastic #ACL2025 workshop on LLM agents, spanning not one but two overflow rooms!! 🔥🔥🔥

It was a pleasure to talk about progress and challenges on the path towards agents that can autonomously make scientific discoveries. I discussed some of our own work on MLGym, HypoGen, LLM-First Search and more. TL;DR: there is plenty of work ahead of us.

🚀 The REALM workshop at #ACL2025NLP was an absolute blast 🔥in the beautiful Vienna 🇦🇹🎻 Kicked off with a packed room. Thanks to the ACL organizers, the workshop was broadcast live to 2 larger overflow rooms and even into the hallway. In total, we had ~400 attendees including

I had lots of fun giving a full-day lecture (+hands-on coding session) at the Armenia LLM Summer School about LLM reasoning and advanced inference-time strategies! 💪 🧠 🔥 Only stayed 2 days but was seriously impressed! so much energy, people super eager to learn and invest in

Congrats to the one and only Yejin!! Always an inspiration!

Internship ServiceNow Research to build the next generation of computer use agents that are safe and secure from malicious attacks. Focus on intervention strategies, defenses to make agents robust against unsafe behavior.. Apply here: bit.ly/3V3mmTg