Even as a computer science / AI researcher, Juliet's topic is super cool. Ant research makes real impact. I am actively engaged in research inspired by ants! There are also many algorithms out there inspired by ants and other animals. Congratulations Dr Juliet Turner !

🚨 Coming up - Research Connections event on Wednesday, November 26th! Are you interested in building interpretable, trustworthy language models and user-centric AI? Or learning about the biases, safety and cultural alignment of language models? Come meet Abhilasha Ravichander and

Share Your Research with Our Open Science Community! 🎙️ Our open science community thrives on collaboration and knowledge-sharing in the various fields of machine learning. We’d love to hear from you if: 🔬You've conducted some exciting research and want to share your findings

There are researchers out there with a unique writing voice. They're actually not that many as far as I can tell. To those people gems, and you probably know yourselves, please please don't LLM it away. Your long time fans will notice and feel a big sense of loss 🥲❤️

Grateful to have met kind souls along my journey in research. I'd protect those people any day, with all my might, and every fiber of my being. Some took big hits to their careers because they did the right thing. If you know folks like that, tell them and protect them.

In 2025, our Open Science Community Leads showed what’s possible when AI research is built in the open. 38 leads, 17 programs, 125 guest speakers advancing open, collaborative AI across the world (find all talks here! cohere.link/uw1lvQQ). 🤯

Mayank Bhaskar Surya Guthikonda Nahid Alam Muhammad Rafi Sudrajat Andrej Jovanović Kanwal Mehreen Manasvi dawane Cole Harrison Leema Krishna Murali Oumayma Essarhi Benedict Emoe-kabu Abrar Rahman - d/acc Bronson Siddhant Gupta Kabir Vivek Silimkhan Esra'a Saleh Rafay neel Arya Suneesh Prahitha Movva @NeurIPS2025 Ujan Anas Zafar Ram Kadiyala Ahmad Mustafa Anis harsha Akanksha @eclecticruchira @tolu_aa va4az Alif Munim (d/acc) Gusti T. Winata Rahul If you believe science works best when it’s open, collaborative, and community-driven, this is your invitation. Explore the programs, join the work, and help build what comes next: labscommunity.cohere.com/community-lead…

In differentiable programs or world models—precision, tolerances, and reductions aren’t trivia—they change control flow and therefore the gradient field you’re attempting to optimize. 0x00b1.github.io/blog/2025/12/2…

this is great! this blog post helps me see two aspects of world models; (1) on a conceptual level we must jump from correlation to causation (literally walking up the ladder,) and (2) on a more grounded level we must consider the (potentially unbounded) length of iterations

Many advancements in AI / ML have been driven by thoughtful small diagnostic tests. Do you know of examples from your research or things you've read? Aiming to highlight this research style in a blog. E.g. Baird's counterexample, the Sussman anomaly, the Rosenbrock function ...

What's the best way to have a non-cloud set up for LLMs to work well with a Latex-focussed writing flow? Open to solutions with a self-owned server. Not looking for an LLM to think with me as much as help with format or fast math writing or locating information in papers.

In the era of pretrained world models, I feel a need to say: Evals only about dynamics prediction do not make sense. We care about *control*. Dynamics prediction evals can look great, but control can be bad. We'll just rediscover the good old objective mismatch problem in MBRL.

once in a lifetime opportunity for you to be the 3000th subscriber to our youtube channel. We have awesome online seminars from some of the leading folks in robotics and robot learning every Thursday at 1pm. cc Glen Berseth

Call for workshop at RL_Conference is open! We are very excited to see your ideas on this year's convergence topics!