🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Update: Code is now available! x.com/_ddjohnson/sta…

Code-change reviews are a critical and time-consuming part of software development. Learn how we applied recent advances in large sequence models in a real-world setting to automatically resolve code review comments in the day-to-day development workflow →goo.gle/3MRg8lY

✨👩‍💻Honored to have been part of this project, and stay tuned for exciting new upgrades soon! Jacob Austin Chris Gorgolewski Valeriya Kharatyan Danny Tarlow Alex Frömmgen, Peter Choy, Gabby Surita, Kevin Villela, and more (and extended contributions, from the Visual Studio Code + Cider-V teams). 🚀

Very happy to share our work on activating Google's software dev process as an engine for ML-powered dev tools. A multi-year effort from many across Alphabet. Special shout-out to Jacob Austin Pascal Lamblin 4️⃣💉😷 Pierre A Manzagol Dan Zheng & Petros Maniatis. See Jacob's🧵& the blog for more.

Super exciting work by many at Google to push forward ML models that help with many aspects of the software development process. Not just simple code completion, but fixing compilation errors, suggesting test cases, helping w/ code review comments, and all kinds of other things.

At ICML for my first conference since before covid. Looking forward to seeing everybody.

✨👩‍💻 Greetings, fellow shape rotators, theorem provers, and code enthusiasts! We're organizing a Code + Math AI meetup at #ICML2023—6:00pm on Thursday, July 27th. There will be appetizers, great company, and also SET card games (if I can scrounge up a few). 😁 📧 DM for

Thrilled to share that I successfully defended my PhD thesis today (earning an excellent grade)! Huge thanks to my committee members Aishwarya Agrawal, Baishakhi Ray & Laurent Charlin. Immensely grateful to Hugo Larochelle & Danny Tarlow for being the best supervisors one could ask for! #PhDone

The way snow is removed in Montreal

New paper: How can you tell when a model is hallucinating? Let it cheat! An expert doesn't need to cheat, so if your model learns to cheat, there must be something it doesn't know. Our general new approach for measuring uncertainty: arxiv.org/abs/2402.08733

I'm constantly impressed by Daniel's ability to generate (and follow through on) ideas that are creative, intuitive, and rigorously grounded. Very excited by this one!

Excited to share Penzai, a JAX research toolkit from Google DeepMind for building, editing, and visualizing neural networks! Penzai makes it easy to see model internals and lets you inject custom logic anywhere. Check it out on GitHub: github.com/google-deepmin…

This blew me away the first time that Daniel showed it to me, and I'm super happy to see it released for everybody. This kind of hands-on interaction with a model that allows inspection and intervention is so powerful towards developing understandings of complicated models.

Excited to share a new blog on ML-based repair for build errors at Google! We found that automatically repairing build errors in the IDE increases productivity as measured by overall task completion with no detectable negative impact on code safety!

Cline now supports Gemini 2.5 Pro -- here's how you can use it in Cline for free: 🧵

Gemini 2.5 Pro is now available in Windsurf! ✨

Gemini 2.5 Pro is available to all Cursor users! You can enable the full 1M context window if you'd like. We're curious to hear how you think it compares to Sonnet.

Wow we just ran Gemini 2.5 Pro on our evals and it got a new state of the art. Congrats to the Gemini team! Sharing preliminary results here and working on bringing it into Devin:

Our latest Gemini 2.5 Pro update is now in preview. It’s better at coding, reasoning, science + math, shows improved performance across key benchmarks (AIDER Polyglot, GPQA, HLE to name a few), and leads lmarena.ai with a 24pt Elo score jump since the previous version. We also

Had a fun conversation with Connie and Logan that gives a peek at what we've been doing to make Gemini better at coding over the last year or so and how we're thinking about what's ahead