Sign in

Elliot H Smith, PhD

@neurosmith.bsky.social
616 followers 863 following 233 posts

Human neuronal computations during cognition, seizures, brain stimulation, etc… @ Utah. Www.neurosmiths.org

PostsRepliesMedia
Reposted by Elliot H Smith, PhD
Queer in AI @queerinai.com · 08/10/2026
Queer in AI is hiring a postdoctoral researcher or research scientist! Join us and help build queer power over AI and data systems: www.queerinai.com/research-sci...
queerinai.com
Research Scientist Position at Queer in AI — Queer in AI
Queer in AI team is looking for a research lead (similar to a postdoc position). The Postdoctoral Researcher / Research Scientist will lead research at Queer in AI, accelerating research at the inters...
056
Elliot H Smith, PhD @neurosmith.bsky.social · 08/10/2026
ladies and gentlemen, I give you Super Intelligence (*stabs heart*):
030
Reposted by Elliot H Smith, PhD
Shrew @shrewshrew.bsky.social · 06/10/2026
I remember when my best friend and genetic clone Karl Diesseroth was the best man AND maid of honor at my wedding and he said in the toast "when I get a nobel prize it will be because of you shrew, your support and intellectual rigor" and I said "not now Karl I'm trying to play balatro"
3264
Reposted by Elliot H Smith, PhD
Laura Gwilliams @lauragwilliams.bsky.social · 06/10/2026
delighted to share my review article, out now in Trends in Cognitive Sciences! i have been thinking a lot about dynamics lately. this article outlines three key principles - persistence, parallelism, and time-stamped encoding - that support human speech comprehension 🧠✨🌀
07224
Elliot H Smith, PhD @neurosmith.bsky.social · 06/10/2026
Move fast; break brains
010
Elliot H Smith, PhD @neurosmith.bsky.social · 06/10/2026
Is intrinsic timescale of FUS way to slow to ever be useful for consumer BCI?
100
Reposted by Elliot H Smith, PhD
Mark Histed @markhisted.org · 03/10/2026
Killed. NSF’s SBE was killed, in violation of law. The President does not have the power to choose to kill programs at a whim by defunding them. The imperial presidency was a mistake and we are in a separation of powers constitutional crisis.
151452529
Reposted by Elliot H Smith, PhD
Lea Duncker @leaduncker.bsky.social · 30/09/2026
If you are an undergraduate student interested in computational and systems neuroscience research, this is your chance to attend #COSYNE27 !!! Apply here for our undergrad travel grant tinyurl.com/cos27ugtravel deadline is Oct 30 (and if you are not an undergraduate, please help to spread the word!)
tinyurl.com
Cosyne 2027 Undergraduate Travel Grant Program
This is the application form for the Cosyne 2027 Undergraduate Travel Grant Program. The deadline to apply is October 30th 2026.
03529
Reposted by Elliot H Smith, PhD
Kim Weeden @weedenkim.bsky.social · 30/09/2026
Trump's effort to rebrand AI as SI doesn't go far enough. It should be called Super Human Intelligence Technology. We'd get more accurate news headlines, like: "Under Trump, Government Turns to SHIT" or "SHIT Heads Visit White House"
681310353
Reposted by Elliot H Smith, PhD
RLDM @rldmparis2027.bsky.social · 29/09/2026
📣 Save the date! RLDM 2027, the Multi-disciplinary Conference on Reinforcement Learning and Decision Making, is coming to Paris, July 6-9, 2027 🎉 Stay tuned for submission deadlines and more updates on our website rldm.org 👀
07037
Reposted by Elliot H Smith, PhD
University of Utah Neurosurgery @uofuneurosurgery.bsky.social · 24/09/2026
@neurosmith.bsky.social
021
Reposted by Elliot H Smith, PhD
bioRxiv Neuroscience @biorxiv-neursci.bsky.social · 21/09/2026
NeuralRNN: a unified framework for recurrent neural network methods in cognitive neuroscience www.biorxiv.org/content/10.64898/20…
051
Reposted by Elliot H Smith, PhD
Abdul El-Sayed @abdulelsayed.bsky.social · 20/09/2026
This is about one thing and one thing alone. Steering our research dollars for cancer, heart disease, and new infectious diseases to Trump’s political allies and friends.
24948391996
Reposted by Elliot H Smith, PhD
maria ruz @mruz.bsky.social · 18/09/2026
We are hiring a postdoc and a predoc to do research (EEG, fMRI, MVPA, RSA) on expectation and attention at the wpd.ugr.es/~humneuro/ in beautiful Granada @cimcyc.bsky.social ! People with interest + relevant experience please reach us at mruz@ugr.es #NeuroJobs #AcademicJobs #CogNeuro #GranadaSpain
wpd.ugr.es
Human Neuroscience Lab – University of Granada
Our lab is located at the Mind, Brain and Behavior Research Center. Here we have access to several behavioral testing rooms, a 3T Siemens Magnetic Resonance Imaging facility with compatible EEG and EyeLink Eyetracker, high-density EEG recording devices, and a Transcranial Magnetic Stimulation device with a robotic assistant and neuronavigator.​
01919
Reposted by Elliot H Smith, PhD
Ignacio Saez, PhD @ignaciosaezphd.bsky.social · 04/09/2026
📣 We're hiring! The Saez Lab at Mount Sinai is seeking a postdoctoral researcher with experience in intracranial recordings, electrophysiology, BCIs, mood disorders, and/or computational modeling, to explore the neural basis of human cognition. 🗓️ Apply by September 30, 2026. Please share widely!
02223
Reposted by Elliot H Smith, PhD
Vinny Costa @vincentcostaphd.bsky.social · 13/09/2026
Apparently the lesson of the digital sphinx by @tuthill.bsky.social & @bingbrunton.bsky.social has already been forgotten. The fly isn’t “playing Doom”. elifesciences.org/reviewed-pre...
elifesciences.org
The digital sphinx: Can a worm brain control a fly body?
1164
Elliot H Smith, PhD @neurosmith.bsky.social · 17/09/2026
From gut bombs to literal bombs
000
Reposted by Elliot H Smith, PhD
Wouter Kool @wouterkool.bsky.social · 15/09/2026
I expect to recruit a PhD student this cycle! The Control and Decision Making Lab at WashU studies control, effort, and decision making, combining behavioral experiments with modeling and neuroimaging. If that sounds cool, check out our work: cdmlab.wustl.edu. How to apply: psych.washu.edu/apply
cdmlab.wustl.edu
Control and Decision Making Lab
14437
Reposted by Elliot H Smith, PhD
Nature Portfolio @natureportfolio.nature.com · 14/09/2026
A paper in Nature #Neuroscience describes a single brain implant that can simultaneously decode speech and gestures in people with paralysis. The findings represent a step towards brain–computer interfaces that enable more natural communication for people with paralysis. go.nature.com/4Ac8BVt 🧪
go.nature.com
Simultaneous speech and gesture decoding for multimodal communication in paralysis - Nature Neuroscience
Brosler et al. develop a brain–computer interface that simultaneously decodes speech and gestures from a single cortical implant to animate a virtual avatar, providing a step toward more flexible communication for people with paralysis.
0196
Elliot H Smith, PhD @neurosmith.bsky.social · 09/09/2026
That was quick!
001
Reposted by Elliot H Smith, PhD
Ben Hayden @benhayden.bsky.social · 09/09/2026
New paper from the Neurosurgery Research Team at BCM! “Neural geometry in the human hippocampus enables generalization across spatial position and gaze” led by @assiachericoni.bsky.social! 🧵 www.cell.com/neuron/abstr...
cell.com
Neural geometry in the human hippocampus enables generalization across spatial position and gaze
Chericoni et al. show that neurons in the human hippocampus track the positions of multiple agents in a simple pursuit-based video game. They identified place codes for each agent and for gaze. The co...
19636
Reposted by Elliot H Smith, PhD
Robert "The Bobby Yaga" McNees @mcnees.bsky.social · 08/09/2026
Ida Henrietta Hyde was born #OTD in 1857. She invented a micro-electrode capable of stimulating and monitoring electrical activity in cells. Later in this thread there will be an amazing story about an octopus and Kaiser Wilhelm II. 🧪 👩‍🔬 🐙 Image: American Physiological Society
Black and white photo of Ida Henrietta Hyde in a lab or lecturing demonstration surgery. She wears a long black dress with billowy sleeves and her hair pulled up. There are shelves or benches behind her.
414844
Reposted by Elliot H Smith, PhD
caroline cash @cashbrowns.bsky.social · 07/09/2026
Nancy 9/7/26 … Happy Labor Day 🛠️
8 Hours for Work, 8 Hours for Rest, 8 Hours for What We Will. Based on a screenprint by Ricardo Levins Morales after the organizing work done by Robert Owen.
70149193713
Reposted by Elliot H Smith, PhD
Ida Momennejad @neuroai.bsky.social · 04/09/2026
Agreed. Also as a community we need to (re)normalize asking critical questions at talks, conferences, random convos. Asking about scientific rigor sometimes takes a backseat to social considerations, esp when someone's charismatic or overconfident (student or prof). Ask kindly/thoughtfully, but ask.
2305
Elliot H Smith, PhD @neurosmith.bsky.social · 04/09/2026
NBD! 👀 👀
000
Elliot H Smith, PhD @neurosmith.bsky.social · 04/09/2026
But what are the brain mechanisms?? 🤔🤔🧠🧠
000
Reposted by Elliot H Smith, PhD
Dan Goodman @neural-reckoning.org · 03/09/2026
#SpikingNeuralNetwork people - this year's SNUFA workshop is live! Talks by: ⭐ Eugene Izhikevich ⭐ @giuliadangelo.bsky.social ⭐ Mihai Petrovici ⭐ Susanne Schreiber Submit your abstracts by Sept 25th. Registration is open! More info at: snufa.net/2026/ cc @fzenke.bsky.social @albada.bsky.social
Picture of the SNUFA spiky hedgehog
02416
Reposted by Elliot H Smith, PhD
Jim Heys @jim-heys.bsky.social · 03/09/2026
I’m very excited to share our new paper, led by the very talented Jack Bowler, a postdoc in the lab! www.nature.com/articles/s41... We all know that the way we learn something matters. But how does a training curriculum actually shape how the brain learns to solve a problem? a 🧵 1/N
nature.com
Structured experience shapes strategy learning and neural dynamics in the medial entorhinal cortex - Nature Neuroscience
The authors use recurrent neural networks to predict how structured prior experience shapes neural dynamics as mice learn flexible timing strategies during complex context-dependent behavior.
47723
Reposted by Elliot H Smith, PhD
The Transmitter @thetransmitter.bsky.social · 02/09/2026
In this episode of @braininspired.bsky.social, @dlevenstein.bsky.social explains how internally generated brain activity during sleep shapes our cognition, and why NeuroAI is such a powerful approach to understanding our brains. #neuroskyence www.thetransmitter.org/brain-inspir...
thetransmitter.org
What’s the role of spontaneous neural activity during sleep?
Daniel Levenstein explains how internally generated brain activity during sleep shapes our cognition, and why NeuroAI is such a powerful approach.
25319
Elliot H Smith, PhD @neurosmith.bsky.social · 02/09/2026
We lost a future neurosurgeon and neuroscientist to this country’s cruelty this week.
000
Reposted by Elliot H Smith, PhD
Biyu Jade He @biyuhe.bsky.social · 01/09/2026
📣 My lab at NYU currently has two open post-doc positions. We are interested in understanding perception, thinking, and hallucinations. We employ a range of cutting-edge techniques in human neuroscience (e.g., 7T fMRI, MEG, iEEG/DES) and computational approaches (e.g., DNN modeling).
26649
Reposted by Elliot H Smith, PhD
Marc Coutanche @marccoutanche.bsky.social · 01/09/2026
We are recruiting an Assistant Professor in Cog Neuro/Cog Psych to join us at Rice! Please spread the word! apply.interfolio.com/192080
apply.interfolio.com
Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio
24642
Elliot H Smith, PhD @neurosmith.bsky.social · 01/09/2026
💯💯💯 the formulaic posts kill me. It feels like an eerie dystopia where everyone has shirked human communication for some bizarre corporate advertisement-speak. 🤖🗣️
150
Reposted by Elliot H Smith, PhD
Victoria Bosch @initself.bsky.social · 28/08/2026
Are brains and artificial neural networks converging onto universal representations? There is a seductive idea making the rounds in NeuroAI / machine learning: train systems well enough, and they all converge on the same representation of reality (i.e. a unique world model). We have thoughts™ 1/n
cell.com
The Umwelt Representation Hypothesis: rethinking Universality
Recent studies reveal striking representational alignment between artificial neural networks (ANNs) and biological brains, leading to proposals that all sufficiently capable systems converge on univer...
818175
Reposted by Elliot H Smith, PhD
keven-pittsinger.bsky.social @keven-pittsinger.bsky.social · 26/08/2026
Linux turns 35 today. Here's your cake. Enjoy.
721712314
Elliot H Smith, PhD @neurosmith.bsky.social · 26/08/2026
Nearly a dime a dozen!!
000
Reposted by Elliot H Smith, PhD
Konrad Kording @kordinglab.bsky.social · 25/08/2026
Thread. I have yet to see a proper explanation of how LLMs actually work that is targeted to an informed general audience. So here is an explainer that I put together along with Tim @countzerozzz (inspired by @CIFAR_News LMB discussions).
1723060
Elliot H Smith, PhD @neurosmith.bsky.social · 22/08/2026
Thanks!!
000
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
I'll stop there, but check out the paper for all the extra tidbits and, most importantly, controls! Super grateful for the fantastic team on this, including @rhiannon-cowan.bsky.social @niloufarshahdoust.bsky.social @rolstonjohn.bsky.social @bkundu.bsky.social & others! end/
010
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
The explained variance of these predictive representations, and their brain area contributions, supported this conclusion, with 'bimodal' participants' neural ensembles having more robust representations and more contributions from the vmPFC. 16/
110
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
These predictive representations didn't exist in 'explorers,' but showed up after outcomes, suggesting 'explorers' choices were more model-free and 'bimodals' more model-based, though not in the conventional sense of MF/MB. 15/
110
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
The cool thing about these low-dimensional, factorized representations in 'bimodal' participants: they showed up in both active trials, in which participants had to stop inflation themselves, and passive trials, in which participants knew when the balloon would stop itself (but orthogonalized). 14/
100
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
We found remarkably similar trajectories in human neural activity that arced around an ordered balloon size axis, except only in the higher performing 'bimodal' participants. Human balloon size axes oriented similarly to those in the deep RL agents: 'bimodal' -> PC1, 'explorer' -> PC2. 13/
110
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
What about the neural geometries? To answer that question, we looked at nearly 600 well-isolated single neuron recordings from medial frontotemporal regions in the participants during the time period in which they made the inference about how big the balloon would be. 12/
110
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
So we found these interesting dynamical motifs in deep RL agents, but what about in the human brain? Well, we found the same proportions of behavioral phenotypes across 44 participants doing a simpler, 3-balloon color version of BART. 11/
101
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
We measured a predicted balloon size axis in these geometrical motifs using linear regression, which we found aligned with PC1 in high-performing 'bimodal' agents, and PC2 in cautious 'explorer' agents. Moreover, balloon arcs aligned on the axes according to the agents' behavior. 10/
101
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
What were the neural representations underlying these different behavioral strategies? In the agents' recurrent layers, we found time resolved geometric motifs that corresponded to elements of the agents' inferences: arcs around an axis that PREDICTED balloon size! (red, smaller; yellow, bigger) 9/
111
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
'Bimodal' agents were overall more successful on the BART, despite popping a similar number of balloons. as 'explorers.' 8/
101
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
One type of agent cautiously explored each balloon size in order to find the maximum size at which to stop balloon inflation. Another just gave up on smaller balloons, in order to leverage the potential of larger balloons. we called these 'explorer' and 'bimodal' agents, respectively. 7/
101
Elliot H Smith, PhD @neurosmith.bsky.social · 21/08/2026
Andy trained 400 agents on a meta-learning version of BART where the agents had to learn when to pop and when to stop 17 different balloon sizes from reward. Then looked at the activity of the agents' recurrent layers, finding they clustered into groups associated with behavioral strategies. 6/
101