It's hard to beat someone at something if they enjoy it and you don't. This is true for good things like writing and dancing and also for bad things like meanness and office politics.

Our paper got an oral at CoRL (rhyme!) We proposed CAMPs, a method for learning to generate abstractions of large planning problems. Stop by at Conference on Robot Learning this week! Paper: arxiv.org/abs/2007.13202 Video: youtube.com/watch?v=wTXt6d… Code: git.io/JTnf6

New blog alert! In our lab blog's first ever post, Leslie Kaelbling writes about the engineering science of embodied intelligence. lis.csail.mit.edu/the-engineerin…

Looking forward to AAAI this week! We’ll be presenting two papers on learning for planning in relational domains...

I recently joined the Structured Intelligence team @DeepMind as a Research Scientist and I'm loving the experience: many interesting, impactful projects and awesome collaborators! If you're a Research Engineer interested in ML4Science and structured ML methods, please apply!

The submission portal for our RSS workshop on learning for TAMP is now open! zt-yang.github.io/rss23-l4tamp-w… The deadline is May 19. We're going to have some really exciting speakers. The format is hybrid, so even if you can't make it to Korea, consider submitting!

If you're interested in LLMs for planning, check out our new preprint: "Generalized Planning in PDDL Domains with Pretrained Large Language Models" arxiv.org/abs/2305.11014 w/ Soham Dan, Kavitha Srinivas, Josh Tenenbaum, Leslie Kaelbling, Michael Katz (MIT-IBM Watson AI Lab)

We are hosting a workshop on Learning Effective Abstraction for Planning at #CoRL2023. Come join us to share insights and discuss the way forward! Website: leap-workshop.github.io Deadline: September 18

Ever heard about "Bilevel Planning" or "Task and Motion Planning", but been unsure what those words mean? Ever wanted a gentle intro to these methods so you can just understand what's going on? Our new blog post might help! lis.csail.mit.edu/bilevel-planni…

나는 과학자라는 직업이 주는 지적인 일상이 좋다. 항상 좋은 지식에 목말라하고, 세상에 필요한 지식이 무엇인지에 대해 고민한다. 내가 창출해낸 지식이 세상에 도움이 되길 희망한다.이런 고민과 목마름, 그리고 희망을 가진 내 일상이 너무 좋다.

파인만 알고리즘의 첫번째 단계는 문제를 적는것이다. 사실 이것은 세상에 필요하고, 또 해결 가능한 문제는 무엇인지를 이미 과학자가 꿰뚫고 있다는 전제하에 가능한것이다. 세상에 필요한것과 해결 가능성을 판단하기 위해선 꾸준히 지식을 쌓아야한다.

아이디어를 떠올리는데 있어서 관련된 글들을 자꾸 읽고 이해하면 그것들이 내 머릿속에서 한데 뭉쳐져서 아이디어가 되는 경험을 자주 하고있다. 따라서 많이 읽은 사람이 결국 좋은 연구를 할수있다. 과학자도 재능보다는 성실함과 부지런함이 답이다.

Great essay by Demis Hassabis comparing AI to the Hubble telescope as an amazing tool for discovery FT Weekend Magazine ft.com/content/048f41…

What distinguishes great founders is not their adherence to some vision, but their humility in the face of the truth.

On average, the deeper people's knowledge of a topic, the more calmly they talk about it.

So a rule that debate be civil doesn't just prevent hurt feelings. It also gives an advantage to those with real knowledge of the subject.

How things get done in the world: Relentless focus Self-confidence Personal connections Effective leadership (communication, management, vision, and evangelism)

Pattern I've noticed in very impactful people: Spend ~1 year exploring broadly, ~4 years relentless focus executing on the most interesting direction, repeat.

Roger Federer gets emotional after his 20th major victory.