🔥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

Can program synthesis learn an abstract world model? Yichao Liang took a big step toward making that happen by using code generation to build a hierarchy of abstractions, grounded in perception and useful for planning.

Lastly, come see the spotlight paper by Jonny So on “Fearless Stochasticity in Expectation Propagation”! It'll be present today at 16:30 in Poster Session 2 East, #4101 arxiv.org/abs/2406.01801

Congrats to my collaborators on this! I think this work is a great example of how deep knowledge of the ML literature (sometimes lacking in the safety community) can help provide insights for AI safety. As Ekdeep Singh mentioned, 1st author @Abhinav37465251 is looking for PhDs!

If you’re planning to go to ICLR, consider submitting and attending AABI! It was an awesome event last year, and this year it will be co-located with ICLR in Singapore!

Now accepted at #ICLR2025! openreview.net/forum?id=QOfsw…)

In case you missed it, we extended the submission deadline for AABI 2025! You’ve got until Feb 14th to send us your workshop and proceedings track submissions! Looking forward to seeing you all in Singapore 🇸🇬

6/13 ICLR submissions accepted! (3 posters, 2 orals, 1 blog post) Congrats to Bruno Mlodozeniec Richard Turner Clement Neo Fazl Barez Neel Alex Shoaib Ahmed Siddiqui Tom Bush Cas (Stephen Casper) Dylan HadfieldMenell and all the other authors! Summaries in thread below... 🧵🧵

Join Us! University Assistant Professor in #MachineLearning #AI Engineering Dept Cambridge University details here: jobs.cam.ac.uk/job/49361/

We’re finally hiring again! Apply to join our group a Professor and you can get to work with some of the smartest students in ML!

🌦️ New AI model aims to provide weather forecasts entirely driven by AI Research published today in nature looks at how Aardvark Weather is reimagining weather prediction. Read our news story: bit.ly/4iG5ThD

The Aardvark nature paper on end-to-end weather prediction with deep learning featured in the The Guardian and Financial Times today. Huge congratulations to all the members of our group and the collaborators! theguardian.com/technology/202…

How do you identify training data responsible for an image generated by your diffusion model? How could you quantify how much copyrighted works influenced the image? In our ICLR oral paper we propose how to approach such questions scalably with influence functions.

All the MLG papers at #ICLR2025 main conference!

That’s a wrap for CBL at #ICLR2025! We’re now getting ready to barge through the NeurIPS submissions to then see you all again soon at ICML!

Great to be back from Singapore from #ICLR2025, and super excited to have given my first oral presentation on influence functions for diffusion models!

New revision of “BNEM: A Boltzmann Sampler Based on Bootstrapped Noised Energy Matching” 🚀 (1/6) We introduce NEM and BNEM — diffusion&energy-based neural samplers for Boltzmann distributions. 📝 Read our paper at [arxiv.org/pdf/2409.09787]

[📜1/9] Does machine unlearning truly erase data influence? Our new paper reveals a critical insight: 'forgotten' information often isn't gone—it's merely dormant, and easily recovered by fine-tuning on just the retain set.

I will likely be looking for students at the University of Montreal / Mila to start January 2026. The deadline to apply is September 1, 2025. I will share more details later, but wanted to start getting it on people's radar!