Excited to share "Exploiting Structure in Offline Multi-Agent RL: The Benefits of Low Interaction Rank", with my former Meta team (Wenhao Zhan, Yonathan Efroni, Daniel Jiang) in collaboration with FAIR (Scott Fujimoto) & Princeton (Jason Lee), is accepted to ICLR 2025!

💫Accepted to ICLR25! 💫 We investigate a special MARL structure in which agents weakly interact. This, we show, makes MARL much more tractable. Led by Wenhao Zhan in his summer internship + it was a delight working on this, and expect to see cool extensions ahead!

There are multiple postdoc positions available as part of an exciting new AI-agent initiative at Columbia that tackles challenges at the frontier of agentic systems and sequential decision-making. I am not very active here so please help me spread the word!

Meet us (but not me sadly) at the poster session: iclr.cc/virtual/2025/p… #ICLR2025 (Also, much more interesting things to explore in MARL and offline MARL imo)

At ICLR 2025 in Singapore, my co-authors and I presented two papers on RL. Feel free to let us know of any feedback and let me know if you'd like to chat! - openreview.net/forum?id=AOlm4… - openreview.net/forum?id=BfUDZ…

Aligned Multi-Objective Optimization (A-🐮) has been accepted at #ICML2025! 🎉 We explore optimization scenarios where objectives align rather than conflict, introducing new scalable algorithms with theoretical guarantees. #MachineLearning #AIResearch #Optimization #MLCommunity

accepted to #ICML25🍁 we asked 🤔 how can we improve gradient-descent in the presence of multiple aligned or similar objectives?🤔 this becomes increasingly important when having access to multiple reward functions / datasets / tasks

we actually started by asking this question in the multi-armed / tabular RL, and after spending some time on it realized it has been explored already by Chris Dann, Yishay Mansour, Mehryar Mohri: proceedings.mlr.press/v202/dann23a.h…

Excited to share that the paper I worked with my previous team on at Meta was accepted to #ICML2025 ! It’s important to explore multi-objective optimization especially in the aligned goal setting. This can be particularly useful in the LLM and agent context where the models are

CATransformers is a carbon-driven neural architecture and system hardware co-design framework. Using CATransformers, we discover greener CLIP models that achieve an average of 9.1% reduction potential in total lifecycle carbon emissions while maintaining accuracy (or increasing

What does it mean to write and think with AI? What new possibilities and challenges does that bring? I spoke with THE AI (in Korean) about our group's research and the future of writing with AI. 👩🤖✍️ newstheai.com/news/articleVi…

Wrapped up Stanford CS336 (Language Models from Scratch), taught with an amazing team Tatsunori Hashimoto Marcel Rød Neil Band Rohith Kuditipudi. Researchers are becoming detached from the technical details of how LMs work. In CS336, we try to fix that by having students build everything:

Use the same Expected Parrot: Open-source tools for AI research survey you ran w/ LLMs with actual humans via our Prolific integration (or bring your own panel) - it works!

Honored that our RL_Conference paper won the Outstanding Paper Award on Empirical Reinforcement Learning Research! 📜Mitigating Suboptimality of Deterministic Policy Gradients in Complex Q-Functions 📎openreview.net/forum?id=H3jcT… Grateful to my advisors Joseph Lim and Erdem Bıyık!