🔥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

Self-Adapting Improvement Loops for Robotic Learning. arxiv.org/abs/2506.06658

Congrats Hao Su Saining Xie for winning the Young Researcher Award! 🎉🎉 Very well deserved. Proud to have closely worked with you both!

wait, speaking of false dichotomies---during your phd, you *can* write code, dive into data and systems, collaborate with a team, and build useful things---all while enjoying complete openness and the freedom to pursue what *genuinely* excites you.

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:

From about 2013-2022, the highest impact thing you could do for AI in the tech industry was publish in academic venues. You didn't have to choose between climbing the ladder and doing open science. Now that world is gone, and I'm still not sure how to navigate this new world.

My Oxford Schmidt AI keynote lecture: How current AI works (its a cultural technology like writing and print), and how it could work to be genuinely intelligent like kids (with intrinsic rewards like empowerment, causal model-building, and care). youtube.com/watch?v=bSlqS4…

We’re bringing powerful AI directly onto robots with Gemini Robotics On-Device. 🤖 It’s our first vision-language-action model to help make robots faster, highly efficient, and adaptable to new tasks and environments - without needing a constant internet connection. 🧵

A mental model I find useful: all data acquisition (web scrapes, synthetic data, RL rollouts, etc.) is really an exploration problem 🔍. This perspective has some interesting implications for where AI is heading. Wrote down some thoughts: yidingjiang.github.io/blog/post/expl…

Now in Nature! 🚀 Our method learns a controllable 3D model of any robot from vision, enabling single-camera closed-loop control at test time! This includes robots previously uncontrollable, soft, and bio-inspired, potentially lowering the barrier of entry to automation! Paper:

Clearing out my old office and finding transparencies from my talks from 25 years ago...

A gentle reminder that TMLR is a great journal that allows you to submit your papers when they are ready rather than rushing to meet conference deadlines. The review process is fast, there are no artificial acceptance rates, and you have more space to present your ideas in the

🎈 Out now: 🎈 "The capacity limits of moving objects in the imagination" (by Balaban & me) of interest to people thinking about the imagination, intuitive physics, mental simulation, capacity limits, and more nature.com/articles/s4146…

When two automated systems ("agents") interact, they must agree on the meaning of the symbols (json, English) that they exchange. People encounter the same problem. What discipline studies how people do this? Would love pointers to the literature 🙏

Papers are reducing to content, you're now fighting the #1 rule: write something people want to read. If you're writing the paper because you need to do a write up and it's a slog of process, turns out your direction probably wasn't impactful. It happens, but learn from it.

Tried Grok 4 on a dozen non-trivial math (under/)grad level math problems. So far, it has failed to fail me even once. Congrats to Yuhuai (Tony) Wu, Eric Zelikman and the whole xAI reasoning team, their progress has exceeded all my expectation!

What exactly is a "world model"? And what limits existing video generation models from being true world models? In my new blog post, I argue that a true video world model must be causal, interactive, persistent, real-time, and physical accurate. xunhuang.me/blogs/world_mo…

yes

There are many great researchers out there. But the ones that really stand out to me are the ones who are also kind, even when they don't need to be.

I think there should be an official satellite location in China, given the fact that a huge amount of NeurIPS works come from China, and so many great Chinese researchers couldn't attend the conference due to the US/Canada Visa issue or long travel distance.

An advanced version of Gemini with Deep Think has officially achieved gold medal-level performance at the International Mathematical Olympiad. 🥇 It solved 5️⃣ out of 6️⃣ exceptionally difficult problems, involving algebra, combinatorics, geometry and number theory. Here’s how 🧵