I’ll be presenting my paper on “Robust Subtask Learning for Compositional Generalization” at tomorrow’s morning poster session at ICML. Consider stopping by if you’re attending!
How can compositionality, a key idea in scaling formal methods, help RL for long horizon tasks? Check out paper with @KeysShore and @obastani to be presented at @icmlconf
Congratulations to fresh @CIS_Penn PhDs @KeysShore and @AalokDThakkar for great dissertations on “specification guided reinforcement learning” and “example guided query synthesis” respectively
Also, thanks to all the collaborators who made my dissertation research possible: @sugumanbansal, @obastani, @ShaanVaidya, Radoslav Ivanav and Steve Hsu.
I successfully defended my thesis today 🥳 Many thanks to my advisor @RajeevAlur for being incredibly supportive throughout my PhD and special thanks to my committee @pappasg69, @thenzinger, @obastani, and Sampath Kannan!
@shortstein Out of the two, using O() seems better because then you can interpret the statement as “there exists an f in O() such that if n > f blah holds”. Omega doesn’t make sense to me because I cannot think of a way to formally and meaningfully interpret n > Omega().
I’m on the job market this year! My work aims to improve ML tools and techniques using ideas from Formal Methods. In particular, I have worked on Specification-Guided RL and Compositional Learning and Verification. More about my work on my website: https://t.co/4KQ3m3s2SY
Looking for Ph.D. students at the intersection of Formal Methods and Artificial Intelligence. Research includes cutting-edge problems on Formal Methods in Reinforcement Learning and classical problems in verification and synthesis. Apply to @gatech_scs by December 15 and DM me!
Prospective PhD students: Interested in doing research on programming languages and systems for Safe, Explainable, and Trustworthy AI? I and other faculty in related areas @CIS_Penn and @PennAsset are hiring! RTs much appreciated!🙏 Read on ... (1/4)
Presenting my work on “Specification-Guided Learning of Nash Equilibria with High Social Welfare” at CAV tomorrow at 2PM. This is joint work with @sugumanbansal@obastani and @RajeevAlur. See you there if you are around!
Paper: https://t.co/mCWpUggYJa
[3/n] Presenting our new work on "A Framework for Transforming Speicifcations in Reinforcement Learning" with @RajeevAlur, @obastani, and @KeysShore at SYNT 2022.
Paper: https://t.co/lSFh5BXssr
[2/n] Presenting our @confCAV 2022 paper titled "Specification-Guided Learning of Nash Equilibria with High Social Welfare" with @KeysShore, @obastani, and @RajeevAlur.
Paper: https://t.co/A7NCXIU8bZ
Poster: https://t.co/bWdnEufjCn
I turned 56 today. I’ve been reflecting on what (if anything) I have learned in my life that I can tell you that might be of use or interest to you.
I’ll spare you the details of the specifics of my history and background and how I got from a poor rural family in India to
PhD update: I have successfully defended my thesis proposal. Many thanks to my committee @obastani, @pappasg69, @thenzinger, Sampath Kannan and my advisor @RajeevAlur!