Your mistakes don't define who you are. It's what you do after you have made the mistakes that makes all the difference. Every mistake you make is a learning experience. They don't make you less capable. But it's how you correct them or learn from them that defines you.

8:00 in Montréal, 14:00 in Málaga, 21:00 in Tokyo and here we are, finalizing details for MODELS Conference that starts next week! Thanks Sébastien Mosser for sharing with us your experience as student volunteer chair of MODELS 2020! cc Lei Ma #models21

Dropping on 22 March 2022.

Today, raia hadsell, Kyunghyun Cho and I are happy to announce the creation of a new journal: Transaction on Machine Learning Research (TMLR) Learn more in our post: medium.com/@hugo_larochel…

2022 年国際女性デーを祝おう #GoogleDoodle g.co/doodle/5n8npf4

Me: Hi yes I’d like a large coffee please Europe:

:-)

Our lab members Da and Zhijie will be presenting our latest work on interactive OoD data detection at #CHI2023. Don't miss the video and connect with them during the event for more discussion. This is joint work with Tianyi Tianyi Zhang from Purdue University.

Our lab member Zhijie Zhijie Wang will present our work on interactive ML model debugging at #CHI2023 , please check out the video and hook up with Zhijie during the conference for further discussion :-). This is joint work with Tianyi Tianyi Zhang from Purdue University :-)

Thanks a lot for the invitation and look forward to joining GAS 2023✨

Here are pictures from the two workshops I participated to at ICSE . Some of my favorite people were there. With hindsight, our community is not only much more diverse than what it used to be when I started, it is also much more friendly and pleasant to attend these events.

The deadline for SE4SafeML (Workshop on Dependability and Trustworthiness of Safety-Critical Systems with Machine Learned Components) is July 10th! Submit your work for presentation to the community. More details on the workshop's website: buff.ly/46gERI7

“Ten years ago I was kicked out and forced to retire.” Our new medicine laureate Katalin Karikó (Katalin Kariko) told us how much it means to be awarded the Nobel Prize after a scientific career that has been full of challenges. Ten years ago, Karikó was still doing all her

The paper "Large Language Models for Software Engineering: Survey and Open Problems" is now on arXiv: arxiv.org/abs/2310.03533 Any comments and suggestions are welcome!!

Stunning results! 🥳 It reminds me of the StyleGAN moment for faces back in 2018. This time, for human generation!

Our new work (arxiv.org/abs/2310.07064) shows that LLMs can learn (sometimes uncommon) rules with 2 stages: (1) induction: generate and verify rules from exemplars; (2) deduction: utilize the rule library for new problems. 11-27% gain on reasoning tasks that require rule learning.

Jailbreaking LLMs typically requires human creativity or hours of compute. Introducing PAIR: a procedure for generating interpretable jailbreaks with only *black box access*, that often succeeds under a minute and only 20 queries! 🧵

I can’t stop watching this. x.com/naminori_bum/s…

Just arrived for this year’s #CHI2024 ! I will present our work “PromptCharm: Text-to-Image Generation through Multi-modal Prompting and Refinement” at 2:45 PM on Monday in 318B. Paper: doi.org/10.1145/361390… Code: github.com/ma-labo/Prompt… (1/6)