CS, psych, and neuro prof @ Stanford. NeuroAI and "regular AI". Also pianos, harpsichords, pipe organs, and bonsai. https://nitter.cf/t.co/xCFbmgSz48
Stanford, CA
Joined April 2009
- Tweets765
- Following651
- Followers5.5K
- Likes921
Pinned Tweet
Streams of thought for natural reasoning in the physical world. *This is PSI-0.6 release* Congrats @khai_loong_aw @KlemenKotar at the PSI team!
Daniel Yamins retweeted
1/ Animals explore without rewards, tasks, or instructions. What computational objective drives that autonomy?
We trained virtual flies with intrinsic-motivation algorithms to see which best reproduce real fly behavior and brain activity.
Work with @rdkeller and @aran_nayebi !🧵
Daniel Yamins retweeted
How do children learn the meanings of their first words? One idea is cross-situational learning: tracking which words go with which things across many situations. In a new preprint, we test this idea at scale on video from children's own perspective 🧵 arxiv.org/abs/2610.31842
ALT Figure 1: Training pipeline. (A) BabyView head-camera videos from 48 children are transcribed and turned into frame–utterance pairs. (B) A vision–language model learns to match words to regions of each frame, and is tested on 4-alternative picture choice trials. (C) Each pair is annotated for whether the utterance refers to something visible in the frame.
Daniel Yamins retweeted
Super excited to share that I'm co-founding Read/Write Labs for #NeuroAI applications to human brain health and human-aligned AI. We could not ask for a more wonderful team: prn.to/4yK9vqo
Daniel Yamins retweeted
Introducing the Precision Brain Atlas: high-resolution neural recordings from 70 patients across 12 academic medical centers. A growing resource for studying human neural function- and a step toward a Whole-Brain Interface.
People are prophecying the death of math but actually in the last few months is the first time that I have even started thinking about hiring a math PhD as a postdoc, because the prospects of doing advanced math in my area are so much brighter now.
Daniel Yamins retweeted
We’ve all seen the cool “Astra for robotics” demos. But behind the scenes? Slow rollouts, hand-engineered APIs, and/or simulated behaviors that struggle on real robots.
How do we make coding agent actually work in the physical world? 🦾 Introducing SimEX: Simulation-Integrated Robotics AutoResearch 🧵
arxiv.org/abs/2609.38982
Daniel Yamins retweeted
Another update from our topography work from (high-schooler!) Navneet Prakash. Navneet trained path-integrating RNNs with TopoLoss and found that they recover a much more brain-like center-surround RF connectivity structure than baseline models (w/o any post-hoc sorting)!
Daniel Yamins retweeted
...that was supposed to read "model-brain alignment is now strong enough" 🤦 Posting early morning on the way to the airport took a toll
Daniel Yamins retweeted
Questions in Systems Neuroscience are indeed changing, but to me the new ones are even more exciting! Some years ago, @JamesJDiCarlo showed an old clip of Hans-Lukas Teuber (the founder of @mitbrainandcog: bcs.mit.edu/events/hans-luka…) saying that science goes through stages: explain, predict, control. #NeuroAI first moved us to predict, and I think it's super exciting that we're firmly moving into control:
- prediction is pretty well established by now (e.g. Brain-Score.org; algonautsproject.com), although models are not perfect and can fail on simple stimuli (what is missing?)
- the field is actually making progress on a theory for why this works (@dyamins & @aran_nayebi arxiv.org/abs/2607.08561, @tyrell_turing nature.com/articles/s41593-0…).
- model-synthesized stimuli can neuromodulate visual cortex and the human language network (@PouyaBashivan @KohitijKar science.org/doi/10.1126/scie…, Walker @sinzlab pubmed.ncbi.nlm.nih.gov/3168…, @GretaTuckute nature.com/articles/s41562-0…, and more)
- even for invasive techniques, topographic models predict and steer the behavioral effects of microstimulation; optogenetics and muscimol (proceedings.iclr.cc/paper_fi…, biorxiv.org/content/10.1101/…)
- and finally model-brain alignment is not strong enough that we can model atypical processing in dyslexia, autism, aphasia, and anhedonia (@melikahnd_1 arxiv.org/abs/2509.24597, pubmed.ncbi.nlm.nih.gov/3570…, arxiv.org/abs/2607.11621, @melikahnd_1 anhedonic-ai.epfl.ch)
How processing breaks down, and how we can push it back toward healthy, are very much algorithmic questions about the brain with a clear real-world impact. And now we can actually test them, how cool is that!
Yeah seriously @SakanaAILabs ImageNet or it didn’t happen… my guess is the true state of the art in bioplausible learning surpassed this years ago …
Hmm, these results are only on MNIST, which even trivial methods like Feedback Alignment work well on. Of course, everything works on MNIST! 😉
Which is why we focused on bioplausible learning that scaled w/ backprop on ImageNet 6 years ago... arxiv.org/abs/2003.01513
Daniel Yamins retweeted
CMU is hiring an Assistant Professor in NeuroAI!
Come join our awesome NeuroAI community 🧠🤖: apply.interfolio.com/189268
Daniel Yamins retweeted
With all the Astra demos, there might be a feeling in robotics of a bitter lesson taking hold?
I think this moment is rather the opposite.
Astra works *because* it leverages traditional human-engineered methods as its tools.
(...for now)
Daniel Yamins retweeted
Hiring postdocs @EPFL: please reach out and/or apply to the AI Center fellowship epfl.ch/research/funding/epf…. My #NeuroAI group is excited about testing and building models of brain and behavior, from sensory through language to cognition, and from healthy to dysfunction (treatment)!
Daniel Yamins retweeted
To learn how brains compute, we need to experimentally adjudicate among competing computational hypotheses. How do we do this in the age of complex neural network models? New review paper: “Making models disagree to learn how brains compute” with @TalGolanNeuro, @SchuettHeiko
My latest substack post describes detailed empirical results for RSA vs ridge regression across a wide range of visual system tests.
I think we're genuinely finally building a better sense of how all these things are related.
Substack: danyamins.substack.com/p/con…
We're testing the consequences of RSA contravariance theory: danyamins.substack.com/p/con…
TL;DR:
- RSA and ridge regression are actually pretty well correlated, especially in high-quality fMRI data.
- The theoretically predicted Ridge-RSA formula works decently!
- BUT: strong behavioral tests expose issues with raw RSA. Ridge model-brain similarity is uniformly more predictive of behavioral match across MANY metrics.
- And the same is true for low-level visual phenomena: Ridge-based model-brain similarity better predicts 20+ V1 physiology metrics.
enjoy!
The Contravariance theory for RSA (and CKA) is here.
Substack: danyamins.substack.com/p/con…
Gory mathematical details: arxiv.org/abs/2607.08561
TL;DR:
- Theiss @ScienceLukas @ermgrant @SaxeLab show RSA decomposes into a "task relevant core geometry" part and a "task irrelevant" nuisance component. The nuisances can dominate the RSA badly.
- The nuisances can be removed by projecting onto the privileged axes. Once you do that, full contravariance applies to the task-relevant core geometry RSA.
- In the hard-task condition, the three metrics (privileged axes, linear regression, and task-relevant RSA) all line up.
Daniel Yamins retweeted
Come work with us! Kempner fellowship is a great opportunity to do exciting research on neuroAI, ML, and theory.
The application is now open for our #KempnerInstitute Research #Fellowship! Postdocs studying the foundations of #intelligence or applications of #AI are encouraged to apply. Learn more and apply by Oct. 1: bit.ly/4pRzHNb #LLMs #NeuroAI #ML
@ShamKakade6 @blsabatini
