Blake Richards

What kinds of support do early-career researchers need? Help The Transmitter and Neuromatch bolster the next generation of neuroscientists.

We want to hear from you! @thetransmitter.bsky.social and @neuromatch.bsky.social are teaming up to better understand how to support early-career researchers. Make sure your voice is heard.

#neuroskyence

www.thetransmitter.org/early-career...

Hey! New faculty! Your annual reminder to keep track of *everything* you do in some way or another. I just dump stuff into a Google doc, nothing fancy, but saves a ton of time when working on promotion & tenure docs.

(Ask me what I'm working on right now...!) 🧪

Neural population-based approaches have opened new windows into neural computations and behavior Neural manifold properties can help us understand how animal brains deal with complex information, execute flexible behaviors and reuse common computations.

Neural manifold properties can help us understand how animal brains deal with competing and multifaceted information, execute flexible behaviors and reuse common computations, writes @mattperich.bsky.social.

#neuroskyence

www.thetransmitter.org/neural-dynam...

Low-hanging fruit, really... 😅

Coming March 17, 2026!
Just got my advance copy of Emergence — a memoir about growing up in group homes and somehow ending up in neuroscience and AI. It’s personal, it’s scientific, and it’s been a wild thing to write. Grateful and excited to share it soon.

📰 I really enjoyed writing this article with @thetransmitter.bsky.social! In it, I summarize parts of our recent perspective article on neural manifolds (www.nature.com/articles/s41...), with a focus on highlighting just a few cool insights into the brain we've already seen at the population level.

A landmark volume, The Handbook of Dopamine, is now online:
www.sciencedirect.com/handbook/han...

Big kudos to the editors, Stephanie Cragg and Mark Walton, for putting this together.

Open-Source Software Engineer As a software engineer on the Thousand Brains Project, you will shape our open-source code base to make it accessible to a large research community. Additionally, you will write code and maintain infr...

🚨 We’re hiring!
Come join the team building a neocortex-based AI system.

Shape the future of AI and robotics.

🔗 Apply here: thousandbrains.org/company/care...

#AI #Neuroscience #EngineeringJobs #ThousandBrains

Honestly, I need to go through the paper more carefully, but I think it highly unlikely that this model will actually stand toe-to-toe against the Geminis of the world on a wide range of non-toy, real-world tasks.

Precision Neurotechnologies Backed by £69m, this programme aims to unlock new methods to interface with the human brain at the cellular level, with unprecedented precision.

🚨 Upcoming postdoc openings 🚨

We will soon be hiring postdocs to work on neural data analysis & modelling in the context of a recently awarded ARIA grant aimed at "precision control" of neural population dynamics (w. collabs. across the UK)

Reach out if interested!

www.aria.org.uk/opportunity-...

🥲

🚨New paper🚨

Neural manifolds went from a niche-y word to an ubiquitous term in systems neuro thanks to many interesting findings across fields. But like with any emerging term, people use it very differently.

Here, we clarify our take on the term, and review key findings & challenges rdcu.be/ex8hW

Persona Vectors: Monitoring and Controlling Character Traits in Language Models Large language models interact with users through a simulated 'Assistant' persona. While the Assistant is typically trained to be helpful, harmless, and honest, it sometimes deviates from these ideals...

Persona Vectors

brb 👀👀👀👀👀👀

Anthropic just dropped this paper. They can steer models quite effectively, and even detect training data that elicits a certain (e.g. evil) persona

arxiv.org/abs/2507.21509

YES! This is exactly what happens to me too. WTF is that, I wonder?

Does anyone else find that flights to NYC are more likely to be delayed than flights to other cities? I dunno if I'm just unlucky, but I think all four of my last trips to NYC have had the flight delayed...

A diagram showing 128 neural network architectures.

How does the structure of a neural circuit shape its function?

@neuralreckoning.bsky.social & I explore this in our new preprint:

doi.org/10.1101/2025...

🤖🧠🧪

🧵1/9

I just wanted to note that the mouse paper was actually the more relevant citation-wise than the "move your head around one" you first cited, since it also showed distinct functional pathways coming from a single objective.

I agree with all this! It's really cool to see what the more performant models achieve here, and as you note, there are several novelties, as you put it! It's a beautiful paper, congrats.

Then, no, unfortunately, I don't...

I mean, technically many countries in Europe would count, but they had to go through that whole WWII and holocaust thing to get there.

Let's hope the process can be less violent this time... 🤞

🔥

Ah, but they cite the wrong paper I've noticed! I let them know... 🙂

Notably, aspects of your results are arguably a replication of that paper, so that's the far more relevant work to cite, @mschrimpf.bsky.social and @davidtyt.bsky.social. 🙂

But, lovely to see you guys take this farther, and in particular, show the more general results in human cortex. ❤️

Very cool work!

One note, we (me, @shahabbakht.bsky.social, @patrickmineault.bsky.social, and Chris Pack) also found that a single, self-supervised loss led to dorsal-like and ventral-like representations in ANNs:

proceedings.neurips.cc/paper_files/...

Many-Two-One: Diverse Representations Across Visual Pathways Emerge from A Single Objective How the human brain supports diverse behaviours has been debated for decades. The canonical view divides visual processing into distinct "what" and "where/how" streams – however, their origin and inde...

🧠 NEW PREPRINT
Many-Two-One: Diverse Representations Across Visual Pathways Emerge from A Single Objective
www.biorxiv.org/content/10.1...

What makes visual processing in the brain so powerful and flexible? Very excited to share our new work where we started from SOTA models that accurately predict dynamic brain activity during hours of video watching, and investigated core computations underlying visual perception

Wow, yeah, that's super exciting to see that result replicated! Very cool study in general too!

A few years ago, with @tyrellturing.bsky.social @patrickmineault.bsky.social and others, we showed one single objective (prediction) was sufficient for explaining specialized pathways of mouse visual cortex: tinyurl.com/2vsjbve7

Super cool to see the same observation now for human visual cortex.

Dopaminergic action prediction errors serve as a value-free teaching signal - Nature Dopaminergic action prediction error signals are used by mice as a value-free teaching signal to reinforce stable sound–action associations in the tail of the striatum.

This is intetresting, paper argues that dopamine also encodes "action prediction errors", i.e. differences between actions you predict you will make and actions you do, regardless of reward. Could be used to reinforce habits for repetitive voluntary behaviours:

www.nature.com/articles/s41...

🧠📈 🧪

Yup, I basically agree with this!