Patrick Mineault
NeuroAI, vision, open science. NeuroAI researcher at Amaranth Foundation. Previously engineer @ Google, Meta, Mila. Updates from http://neuroai.science
- New post: things have been moving very fast in AI. But has safety caught up to capabilities? open.substack.com/pub/naix/p/w...
- Any good neuroscience/NeuroAI discord servers I should know about? I feel like the neuroscience town square on bsky is a bit of a ghost town
- Loving the new ICLR blogpost track. Here's a cool one on engrams in neuroscience vs mechanistic interpretability: d2jud02ci9yv69.cloudfront.net/2025-04-28-e...
- Can o3 curate your reading stack? I asked it about last week in NeuroAI, and while not great, it is much better than anticipated, highlighting 3 solid talks/papers and several other reasonable suggestions. I wouldn't be surprised if this was solved in a couple of gens chatgpt.com/share/680bae...
- Reposted by Patrick Mineault[Not loaded yet]
- Excellent talk from Zico Kolter at ICLR on AI safety. Science of deep learning and AI safety are some of the most impactful ways academic AI researchers can bend the curve
- Some fantastic work on the virtual fly www.nature.com/articles/s41...
- Reposted by Patrick Mineault[Not loaded yet]
- I asked o3 for help with LaTeX quote boxes. This is the example quote it used. The machines are on to us.
- Crowdsourcing this: what are the best examples of scaling laws in foundation models for neuroscience? My top 3 examples: Ctrl+Labs, Sato et al., Mineault et al. @colehurwitz.bsky.social , @mehdiazabou.bsky.social , @nandahkrishna.bsky.social , @averyryoo.bsky.social, @tyrellturing.bsky.social ?
- Reposted by Patrick Mineault[Not loaded yet]
- The Foundation Model crew at #COSYNE2025. With @josueortc.bsky.social , @colehurwitz.bsky.social , @averyryoo.bsky.social , @nandahkrishna.bsky.social , @auschulz.bsky.social
- Reposted by Patrick Mineault[Not loaded yet]
- This is a really interesting (and heroic!) map of "dark matter" in the sciences that should be addressed with new tools and datasets. What's the dark matter of neuroscience?
- We (mostly @adammarblestone.bsky.social and Mary with the help of the amazing scientists in our network) made a map of “fundamental development” gaps in science and technology. Go explore gap-map.org.
- Scott Linderman and I ended up on the same panel at #ODIN2025 at the @alleninstitute.bsky.social , wearing the same J Crew shirt. Twins!
- Nice article on MICrONS in the NYTimes from @carlzimmer.com www.nytimes.com/2025/04/09/s...
- Headed to Seattle this week for ODIN workshop at the Allen Institute. Hit me up!
- . @kordinglab.bsky.social 's substack is great. Lots of food for thought from one of the most piercing thinkers in NeuroAI. kording.substack.com/p/why-are-we...
- Brain-Inspired Intelligence to Evolutionary, Collaborative, and Safe Systems. This paper looks wild: 264 pages and 1416 references. arxiv.org/abs/2504.01990
- And that's a wrap for Cosyne, safe and sound in NYC! Always happy to meet friends old and new. So many people thanked me for an old paper, a blog post I wrote, the good research guide, Neuromatch... I will hold on to that gratitude, it means a lot!
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- Headed to Mont-Tremblant for COSYNE—looking forward to see everybody!
- Excited to go through this! One day we'll understand how these dang things work
- Reposted by Patrick MineaultFun with worms on @WIRED www.wired.com/story/openwo... We've been thinking about how #celegans can revolutionize science again, together with @kordinglab.bsky.social @eboyden3.bsky.social @wormsense.bsky.social and many other (also find our preprint on #arXiv)
- I am fully aligned with Eva and Blake's vision. I published this piece a couple of weeks ago: www.neuroai.science/p/what-are-f...
- Eva Dyer and I wrote an opinion piece for @thetransmitter.bsky.social on why neuroscience needs to embrace complexity and accept the "bitter lesson" by using a data-driven regime at scale. With commentary from several wonderful researchers! 🧠📈 #NeuroAI 🧪
- I've got a couple of blog posts in various states of drafting: one on my mental model for AI safety and why I'm more worried now than 6 months ago, one on mech interp, and another one on the coordinated research program/FRO/ambitious science ecosystem. Let me know which one you'd like fast-tracked!
- I like to use a diagram on the second slide of my presentations to help people make sense of my unusual background. Unfortunately, I ran out of space with the old one, and 5 circle Venn diagrams are unwieldy. How about this one?
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- Reposted by Patrick Mineault[Not loaded yet]
- How can foundation models accelerate neuroscience? It's not just the models; it's the ecosystem. I went into a journey into the field of synbio to see how machine learning can accelerate science www.neuroai.science/p/what-are-f...
- Super excited to finally get to read this: Ezra Klein and Derek Thompson book-long essay on how and why to build more sci-fi stuff www.simonandschuster.com/books/Abunda...
- Reposted by Patrick Mineault[Not loaded yet]
- Great set of talks on whole-brain sims featuring Philip Shiu, @lappalainenjk.bsky.social and @kordinglab.bsky.social www.youtube.com/watch?v=tebD...
- Not to toot my own horn, but Neuromatch's NeuroAI course is one of the best places to learn about this exciting area of neuroscience. Students and postdocs, apply!
- One trend that I feel hasn't been priced in to our projections of future neuroscience: it's becoming far easier to design new indicators and receptors. A couple of recent examples: www.biorxiv.org/content/10.1...
- Between paperskygest and my custom science feed, my bsky information diet has become much better. Try these bsky.app/profile/pape... bsky.app/profile/patr...
- Reposted by Patrick Mineault[Not loaded yet]
- The new ICML review format is *a lot* for reviewers. I like its thoroughness but I'm not sure the added SNR–if any–is worth the hassle.
