Jens-Bastian Eppler

At the very least, it is completely superfluous and can just be omitted.

How, then, would the same stimulus within one session result reliably in the same neuronal response?

Did you watch the video? Or read the article?
I have the feeling, you're talking about something completely different. We are talking "representational drift". Not movement detection. How would movement detection account for different responses to the same stimulus on different days?

YouTube video by Centre de Recerca Matemàtica Why Your Brain Is Never Still: Representational Drift and Statistical Learning

Direct link to the video 🎥
youtu.be/z63fmYSBcB0

And the excellent article the outreach team wrote for the CRM homepage:
www.crm.cat/why-your-bra...

@crmatematica.bsky.social
🧠🧪

Loved working with our amazing outreach team on this short video about representational drift! @crmatematica.bsky.social

🧠🧪

In it, I explain the points we make in our recent review in CONEUR:

doi.org/10.1016/j.co...

I’ll address this question in a minimal model, which I’ll present at the Bernstein Conference (29 Sep – 02 Oct) in Frankfurt.

If you’re interested, come to our workshop or visit my poster. Looking forward!

@bernsteinneuro.bsky.social
@crmatematica.bsky.social

🧠🧪

YouTube video by The Theoretical Neuroscience Channel Representational drift reflects ongoing balancing of stochastic changes by.... | Jens-Bastian Eppler

My talk at the WWTNS is now online!
In it, I explore how both random processes and Hebbian learning shape representational drift:

www.youtube.com/watch?v=WH4P...

In the end I raise the question:
How can neuronal activities change while representational similarity is preserved?

🧠🧪

This is the first paper to come out from my postdoc at @crmatematica.bsky.social.

Still in collaboration with my old friends Simon Rumpel and Matthias Kaschube, though. 🧠🤝

5/5

This process can be likened to "herding cats": the stochastic changes (cats) are kept in check by statistical learning (shepherd & dogs). We even added an illustration containing 🐱 & 🐶. 🧠🧪

4/5

We propose: ongoing statistical learning is not only for adapting to new environments, but also to continuously maintain memories in the face of stochastic changes.

Without it, the system would lose track. With it, the cognitive map stays intact. 🔄🧠

3/5

This means that what we call “representations” (Edelman, 1998: "Representation as representation of similarities") are actually stable during “representational drift” - even though neuronal activities change. 🧠🔍

A stable percept can thus emerge from an unstable substrate.

2/5

Following up on my last post with some more detail:

Our new opinion in CONEUR: Statistical learning and representational drift: A dynamic substrate for memories

authors.elsevier.com/sd/article/S...

Neural activity patterns drift over days. Yet, representational similarity remains stable. 🧠🧪

1/5

New opinion paper out in CONEUR:
Statistical learning and representational drift: A dynamic substrate for memories

We propose: Representational drift arises from stochastic changes balanced by statistical learning.

Probably with the craziest figure I’ll ever publish…

🧠🧪

doi.org/10.1016/j.co...

I like the freedom to follow my own ideas and availability to get feedback on those ideas.

Late postdoc stage, though. During my PhD I liked more and closer guidance better (in exchange for less freedom).

I hope, it is not always 1:1 the environment. :D If the brain would only reflect the environment, how would we be able to learn at all?

I am really excited about all the evidence on learning and manifolds. So much still to be found and understood.

And for me the big question is: What about representational drift? 😄

Manifolds can be mapped, but individual neuronal responses change. That’s neat, but how does it make sense?

How can perception stay stable if the neurons keep changing?
I’m sure manifolds will help us understand that too…

🧠🧪
3/3

Brain structure and behavior give rise to manifolds.

But what about the environment? Either through statistical learning or just similar inputs leading to similar representations: the input/task will shape the manifold, no?

Linear track → linear representation.
Circular color space → ring.

🧠🧪
2/3

A neural manifold view of the brain - Nature Neuroscience Recent advances in neuroscience have revealed how neural population activity underlying behavior can be well described by topological objects called neural manifolds. Understanding how nature, nurture...

Great new review on neural manifolds by @juangallego.bsky.social and @mattperich.bsky.social.

Amazing read. Lots of intuition and examples. Great synthesis of why manifold structure pops up everywhere in neuroscience, from motor control to cognitive tasks.

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

🧠🧪
1/3

:D

I do think excitability plays a role and I really like the tuning changes through lateral inhibition.

I'm not sure about the significance of excitability, as already minor synaptic weight changes can cause huge response changes. And synaptic weights change considerably.

New preprint by @carolinehaimerl.bsky.social:
A self-consistent model reproducing key features of representational drift without synaptic changes.
Great piece of theoretical work!

That said, I think it remains an academic exercise. Synapses change, all the time.

www.biorxiv.org/content/10.1...

🧠🧪

Representational drift as the consequence of ongoing memory storage - Scientific Reports Scientific Reports - Representational drift as the consequence of ongoing memory storage

Great new paper by my colleagues Gloria & Alex, and @lichengzou.bsky.social. Hebbian plasticity accounts for much of representational drift: Drift is ongoing memory storage, not noise. Also reconciles contradictory findings on stability after repeated exposure.

www.nature.com/articles/s41...
🧠🧪

⏳ Register for the #BernsteinConference by July 30 and secure the early-bird fees!

Register here 👉 bernstein-network.de/bernstein-co...

An alle Forscher*innen, die KI verwenden/erforschen und deutsch sprechen: Macht mit!
Ich war jetzt schon paarmal dabei und es hat immer sehr viel Spaß gemacht.

(Dieses Mal habe ich leider keine Zeit...)

🧪

:D Ich habe das Zeichenlimit bis auf das letzte Zeichen ausgenutzt...

Habe beim deutschen #ImAScientist zum Thema KI mitgemacht. Dort können Schüler*innen mit Wissenschaftler*innen chatten.

Ist oft super chaotisch mit einer ganzen Klasse zu chatten, aber macht richtig Spaß. Kann ich wirklich nur empfehlen!

Meine Eindrücke dazu:
imascientist.de/2025/07/09/d...

💬🧪🧠👩‍🔬👨‍🔬

I'd be happy to, if you still need someone.

It is. :D

Don't forget to submit your abstracts! And see you in Frankfurt. :D

If you still need one, I'd be happy to.

Just submitted my abstract. :)
You still have the whole weekend to submit yours for a contributed talk at the #BernsteinConference 2025!

Probably my favorite conference, and once again the speaker lineup looks fantastic. Hope to see many of you in Frankfurt.