Markus Meister

Mice in a labyrinth show rapid learning, sudden insight, and efficient exploration Mice exploring a labyrinth freely for the first time learn a complex action sequence after a handful of rewards, exhibiting a learning rate 1000-fold higher than in commonly-used paradigms.

Whoa! Robots are catching up to mice. Expert navigation after an hour. Though the mice are not human-guided. And they can do it in the dark.
elifesciences.org/articles/66175
elifesciences.org/articles/84141

I wonder how long they spent revising this...

Technically this is a motorcycle

In a monarchy or dictatorship, the people know they're not responsible for what happens. So rather than arguing with each other, they can band together every so often and throw the bastards out. 2/2

An insidious aspect of democracy is that it makes the voters blame themselves (or other voters) for bad outcomes. This is just victim-shaming. Most voters (me included) know nothing about how to run a country, and their votes can be steered in any desired direction by simple propaganda tools. 1/2

The Maintenance Race - Seventh Draft | Books in Progress Books in Progress is what we call a “public drafting tool”: Drafts will be made available for comment from the public, allowing for direct collaboration between author and reader.

Really enjoyed this story about maintaining your sailboat (and your sanity). "Sometimes maintenance involves shooting the shark." books.worksinprogress.co/book/mainten...

@erictopol.bsky.social posted this diagram earlier today. When I pointed out the obvious AI slop and copious errors, he blocked me. But you can also find it on his substack. Topol poses as a serious medical person, but disseminating such graphical nonsense is disqualifying. Unfollow.

Ugh please save us the AI slop. What are the green spidery thingies? Are their insides connected to the CSF tube? And what's that brain tissue outside the meninges? Did you discuss those issues?

What human intelligence are you talking about? At the intersection of neuroscience and artificial intelligence there is a popular trope that marvels at how our brain can accomplish intelligent functions with just 20 watts of power supply. Where…

That's the minute I spend waiting for ChatGPT's response. See my related rant here: markusmeister.com/2025/09/15/w...

Adam Kampff’s passion for understanding and explaining the world was unmatched. Living by example and not ever compromising on his dreams, Adam was uncanny in making people realize they can learn and understand anything and everything. Keep his dream alive!
In his own words: tinyurl.com/ye29csw3

Suggested title: "Not Even Bonkers"

I'd like to learn more about (1) the mechanisms that create fake jobs so efficiently, and (2) their coexistence with the pressure for productivity. E.g. Amazon: Bezos fires the least productive 15% of his employees every year, yet is happy to spend a fortune on so-called "legal services".

Society has gone through enormous changes in productivity already, and the answer was always to invent pointless jobs, rather than let people go unemployed. Many occupations today only exist as a vehicle to distribute paychecks to the population. See also David Graeber's "Bullshit Jobs".

A multi-institutional investigation of psilocybin’s effects on mouse behavior Studies reporting novel therapeutic effects of psychedelic drugs are rapidly emerging. However, the reproducibility and reliability of these findings could remain uncertain for years. Here, we impleme...

On the other hand, see this interesting null report about psilocybin effects in mice: www.biorxiv.org/content/10.1...

When I was little, and wondering whether to study physics, his book Physics in the Twentieth Century made a big impression. Still a good read today.

Synchronous period-doubling in flicker vision of salamander and man - PubMed Periodic flashes of light have long served to probe the temporal properties of the visual system. Here we show that during rapid flicker of high contrast and intensity the eye reports to the brain only every other flash of light. In this regime, retinal ganglion cells of the salamander fire spikes o …

But maybe not 10 years ago? A while ago I rediscovered a thing, and found it had been discovered every 25 years, going back all the way to Lord Adrian. Come to think of it, it’s time again now… pubmed.ncbi.nlm.nih.gov/9535954/

It’s a rare neuroscience paper that makes a claim that could be proven wrong, even in principle.

Stern 1-2007 | PDF | Experiment | Science The document discusses an interview with Peter Stern, an editor at Science magazine. It covers topics like the submission and review process, changes over Stern's 10 years as an editor, popular fields...

Do you mean papers in Science magazine? I would guess 90% or so. But it's an unfair question, because being right isn't part of their editorial goals. See, for example, Peter Stern in this interview: www.scribd.com/document/382...

Imagine showing a Patek-Philippe to someone who had only known an Apple watch. Eventually combustion vehicles will have that same old-timey precision-engineering appeal.

And - helpfully - many authors just call them P-cells and M-cells.

Also: Parasol ganglion cells start the Magno pathway, while Midget cells start the Parvo pathway.

Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina - PubMed The development of orderly connections in the mammalian visual system depends on action potentials in the optic nerve fibers, even before the retina receives visual input. In particular, it has been suggested that correlated firing of retinal ganglion cells in the same eye directs the segregation of …

I know because I discovered them.
pubmed.ncbi.nlm.nih.gov/2035024/
They are based on conventional synaptic transmission. And fortunately they stop once the retina has to do something useful, like seeing.

I so miss Bob! Any chance for early parole?

Footnote: The retina does contain an "ephaptic synapse" between horizontal cell and cone terminal. The HC injects current into the synaptic cleft that alters the potential across V-dep channels in the cone. The tight geometry constrains the interaction to just those two neurons. Cool biophysics! 8/8

At the other end of the nervous system, the waves and oscillations then hand control back to the spiking motorneurons so they can make the muscles twitch. Like I said: conceivable, but I wouldn't bet on that picture. 7/8

It is conceivable that somewhere on the route from sensory periphery to central brain regions, the spiking neurons hand over the reins to coarse waves and oscillations, and those become dominant and implement cognition. 6/8

So in the periphery, our models of circuit function are quite powerful at explaining things and predicting function, and they don't need LFPs and oscillations. 5/8

A few years ago, Sergio Neuenschwander revisited this work and found that the oscillations were all an artifact of anesthesia. Incidentally, this was a courageous and hugely important thing to do. I'm not sure what the implications are for all the claims about cortical oscillations and binding. 4/8

Local field potentials and oscillations are not part of that model of the retina [but see footnote]. There was a brief scare in the 1990s when Wolf Singer's group claimed that cortical oscillations were the solution to the binding problem, and that they traced all the way back to the retina. 3/8

For example, the retina. A pretty complex circuit, ~140 cell types, uses most of the neurotransmitters known to man, and performs some serious computations on its input. We have a useful model of how the circuit works, that can predict (to ~80%) the spiking output for a new visual stimulus. 2/8