Xinyu Wang

@patrickshafto @TaliaRinger That'd be great! Thanks. Looking forward!!

@xwangsd @TaliaRinger I'll tweet it out shortly

I'm at this DARPA expMath (AI for math) program kickoff and the lineup is legendary. Like, at the level of a truly historical meeting, held at a historical time. Makes me feel way more important than I actually am lol

@ilyasergey @kc_srk @RanjitJhala Expert assisting the tool?

@RanjitJhala @kc_srk Thank you Ranjit! Indeed, we can finally use those nerdy proof tools as, what’s the word… assistants!

New on "Proofs and Intuitions": Verifying Move Borrow Checker in Lean: an Experiment in AI-Assisted PL Metatheory. proofsandintuitions.net/2026/03/18/mov… The gist: I formalised Move's type system in Lean: 39KLOC, under a month, with Claude. Person-years in PL research are now person-weeks.

@WhiterMeerkat @awsTO Happy to chat if you're serious :-)

@awsTO @xwangsd let’s start one

RIP Tony Hoare. 😢 I only met him a couple times many years ago at MSR and was not accomplished enough at the time to have much useful to say, but he sure cast a long shadow over the field...

@krismicinski Good luck!

@swarat Congratulations Swarat!!

Honored to be named a 2025 #ACMFellow! Many thanks to my students, collaborators, and mentors over the years -- working with you has been a privilege. And a big thank-you to my family for their perennial love, support, and patience.

@AI4Code Congratulations!

Penn-goin! Elated that my amazing daughter Isha was admitted to Penn in the Mechanical Engineering program's Class of 2030! Lots of gratitude to her teachers, our family and friends, my colleagues, and even my students who helped in any way possible. We as a family got to experience the stressful process of US college admissions firsthand. Sending good wishes to others going through it right now. A reminder that admission decisions are never a measure of your worth as a person. thedp.com/article/2025/1… Isha's website with artwork and writings: ishanaikfineart.com #PennEngineeringProud @MEAM_Penn #classof2030

@Skiminok @reflection_ai Congrats!!

🎉 Next week, I am excited to join @reflection_ai as a Member of Technical Staff to help build the open intelligence ecosystem of the Western world. It's the most exciting opportunity to help software builders in our time, and will shape many years of AI Engineering in the medium-term before AGI. Not just about Western vs Eastern open models, but more about how AI-driven software will look like in 2030. I spent some time articulating my thoughts about where we're going as a community and why... which became a whole blog post. Take a look, hope it interests you! (And if it really does, we are hiring in NYC, SF, and London 😉) alexpolozov.com/blog/reflectio…

@Skiminok @julesagent All the best, Alex! Look forward to your update on the next stop.

Ending a chapter is always bittersweet. I love Google, I adore DeepMind, and will continue rooting for them in our once-in-a-generation era of building AGI 🚀 To my friends in Gemini, in @julesagent, and in the internal moonshots – you rock and I'll cherish the time working together! And who knows, maybe I'll back again soon 🙂 But it's time to try something different. I'll share more after a small break.

@ccrisccris @fchollet @VictorTaelin Thanks! Just a little bit - and we've started to look into it more! Working on something (which is general and we think is very cool) for SQL, and hoping to apply it to ARC problems. :-)

@xwangsd @fchollet @VictorTaelin Love that. Have you looked into application to ARC puzzles?

Having fun exploring a new take on compositional program synthesis. The goal is to use composable abstractions (and make @fchollet happy), and obtain an exponential reduction of the state space (and make @victortaelin happy). The key insight: repeatedly solve simpler abstractions of the task, then refine back to concrete domains. Think CEGAR but for program synthesis from examples. Instead of attacking the whole search space at once, we lift problems to abstract domains, solve there, then embed back. Branching + voting handles abstraction failures. Two toy instantiations tested on @arcprize and inductive programming (e.g., as in @SumitGulwani's Flash Fill) show meaningful speedups, suggesting the approach is worth exploring further. Feedback welcome! 🧵

@gregd_nlp All the best, Greg!

📢I'm joining NYU (Courant CS + Center for Data Science) starting this fall! I’m excited to connect with new NYU colleagues and keep working on LLM reasoning, reliability, coding, creativity, and more! I’m also looking to build connections in the NYC area more broadly. Please reach out if you're interested in chatting! This move comes after 8 years working with incredible students and collaborators at UT Austin. Thank you to everyone who supported me in my first academic appointment; I look forward to continuing our collaborations but I will miss you! (and the breakfast tacos!)

“How I Became a Storyteller (and How You Can Too)” — the most important talk I’ve ever given. And now, it’s yours to experience. youtu.be/ur_YQpVgW5s

@SharonYixuanLi Congratulations!!!

It’s official: I got my tenure! Immensely grateful to my colleagues, students, friends, and family who have supported me on this journey. On, Wisconsin!

@rishabhs @databricks Congratulations Rishabh!

I'm super excited to share that I recently joined the @databricks AI research team to help with AI for data science efforts. We are working on real-world AGI to help customers succeed on the Databricks platform. We are hiring, please join us in this exciting mission!

Thrilled to join the UMich faculty in 2026! I'll also be recruiting PhD students this upcoming cycle. If you're interested in AI and formal reasoning, consider applying!

@KonsKallas Congratulations!

Were you looking for benchmarks for your shell research? Look no further! Our 🐨 Koala 🐨 benchmark suite is out (github.com/kbensh/koala) and it just got a best paper award at the last ATC ever! Check it out :)

@ilyasergey And you only know that's the case when something false you know of was proved

OH: “It’s okay to assume false axioms as long as you only use them to prove true statements.”