Leading the Cognition, Computation, and Consciousness Lab at Vanderbilt University
Nashville, TN
Joined November 2021
- Tweets464
- Following358
- Followers1.3K
- Likes456
André M. Bastos retweeted
We are offering 10,000 scientists the highest usage tier in Google Antigravity free of charge for 6 months. Please apply! forms.gle/WvmBbPvRbpKG1BPn9
André M. Bastos retweeted
Come join us for the second Neuro for AI and AI for Neuro workshop. Looking forward to seeing YOUR paper submission.
We’re excited to announce the 2nd International Workshop on Neuro for AI & AI for Neuro, taking place Feb 22–23, 2027, in Montreal 🇨🇦 alongside @RealAAAI 2027! 🎉
📢 Call for Papers is now open!
For more info and submission details: neuroai-multimodal-workshop.…
#NeuroAI #AAAI2027
André M. Bastos retweeted
Computation through morphology. Neuronal morphology is key to how brain microcircuits compute: dendrites shape input integration, while axonal arbors determine who connects to whom and where. These architectures organize excitation, inhibition and information flow. #Neuroscience
André M. Bastos retweeted
Meet the dark guardians of movement: dopaminergic neurons in the substantia nigra. Identified by Annica Dahlström, Kjell Fuxe and team in 1964, they supply dopamine to the striatum, helping us move smoothly. Their degeneration is a hallmark of Parkinson’s disease.
André M. Bastos retweeted
Giants in the insula! Von Economo neurons stand tall with their large, spindle-shaped bodies. Described by Constantin von Economo in 1926, these layer V cells also inhabit anterior cingulate cortex. Their role in emotion and social cognition remains an intriguing mystery.
André M. Bastos retweeted
A spider neuron that stutters inhibition in cortical layer I. The neurogliaform cell weaves a dense axonal web around pyramidal cells dendrites, spreading GABAergic inhibition through its local circuit. In 1895 Cajal described this tiny cortical inhibitory cell. Amazing 👇
5/ This was a huge effort:
@VanderbiltU: @jake_westerberg, @SophyXiong, @HNyXJ (BastosLab) recorded the NHP data.
@AllenInstitute: OpenScope recorded the mouse data.
@EliSennesh joined in analyzing this cross-species data.
9/ Now comes the future: w new NIH R01 funding, we’re scaling from 128-channel probes toward 512–1,024+ channels (with @DiagBiochips)
Cortex + thalamus + basal ganglia + hippocampus.
The goal: watch predictions and prediction errors unfold across the distributed circuit.
10/ And we’re hiring a postdoc at Vanderbilt to help build and use this next generation of large-scale NHP neurophysiology.
If this sounds like the kind of neuroscience you want to do, take a look:
vanderbilt.edu/postdoc/prosp…
1/ We’ve spent the last few years building ways to record enough of the brain at once to actually test theories of cortical computation.
Now we have major new funding to hire a postdoc to help us do it.
Here’s a 🧵 (or skip to end for job link!)
4/ But predictive coding is fundamentally a multi-area hypothesis.
So we built MaDeLaNe:
Multi-Area, high-Density, Laminar Neurophysiology.
Together with OpenScope + the @AllenInstitute, we tested predictive coding across cortical areas, layers, cell-types AND species.
3/ Then we started asking what those rhythms do.
Our gamma–beta push-pull model links alpha/beta to predictive control and gamma/spiking to information that escapes prediction.
@SophyXiong tested this by perturbing the system with propofol:
doi.org/10.1073/pnas.2315160…
2/ First came a striking observation.
Across 14 cortical areas, we discovered the “spectrolaminar motif”: gamma/high frequencies strongest in superficial layers; alpha/beta strongest in deep layers.
Mendoza-Halliday et al., Nature Neuroscience:
nature.com/articles/s41593-0…
1/
🧠 Postdoc opening in my lab at Vanderbilt! We're scaling laminar recordings from 128 to 512-1,024+ channels to follow predictions & prediction errors from cortex to thalamus, basal ganglia & hippocampus in behaving macaques. NHP experience not required 🧵
3/
Who we're looking for: PhDs (or soon-to-be) in neuro, comp neuro, neuroengineering, cog neuro, or related fields. Ephys, modeling, signal processing, or neurotech all welcome. Excited about building agentic AI for data analysis? We especially want to hear from you.
4/
The lab is newly R01-funded and growing, with collaborators across Vanderbilt and beyond. And Nashville is a great place to live 🎸and work. Send CV, research statement & references to [email protected]. Rolling review. RTs hugely appreciated 🙏