Lightnews — Scholar-powered news

Reposted by Georg Keller

Kris Jensen @kristorpjensen.bsky.social · 14d

I’m super excited to finally put my recent work with @behrenstimb.bsky.social on bioRxiv, where we develop a new mechanistic theory of how PFC structures adaptive behaviour using attractor dynamics in space and time!

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

9 85 200

Reposted by Georg Keller

FMI science @fmiscience.bsky.social · 21d

This year's Ruth Chiquet Prize goes to @solygamagda.bsky.social, who sent a video message, for her work on how the brain detects sensory mismatches. Read more at: www.fmi.ch/news-events/...

3 11

Reposted by Georg Keller

FMI science @fmiscience.bsky.social · Sep 8

🚨 We're hiring, please share! The FMI seeks a tenure-track Group Leader (Assistant Prof) in Structural Biology 🔬
Innovative scientists in genome regulation, RNA metabolism, or protein homeostasis—especially using cutting-edge approaches—apply now at www.fmi.ch/education-ca...

66 61

Georg Keller @georgkeller.bsky.social · Aug 19

I often get asked “do you think visuomotor mismatch responses are a mouse thing?” – it looks like at least humans have them as well. very excited by this!

Magdalena Solyga @solygamagda.bsky.social · Aug 19

A new preprint from our lab with @zelechowski.bsky.social & @georgkeller.bsky.social !

Using wireless EEG + VR, we recorded visuomotor mismatch responses in freely moving humans.

Huge thanks to all participants, Keller Lab members and FMI facilities!

Read more: www.biorxiv.org/content/10.1...

1 10

Georg Keller @georgkeller.bsky.social · Jul 30

If so, this would mean - as you also propose - that the responses indeed won't always be easily interpretable.

1

Georg Keller @georgkeller.bsky.social · Jul 30

My current best guess is that cortex implements something like a JEPA and uses local error computations to implement credit assignment (or backprop if the system is hierarchical) the way Rafal Bogacz has suggested.

2 3

Georg Keller @georgkeller.bsky.social · Jul 30

I also think the predictive processing model is wrong, but for slightly different reasons. These pertain to the way superficial and deep layers in cortex appear to interact that are not consistent with predictive processing (@loghyr.bsky.social‬ will have a preprint on this soon).

2 1

Georg Keller @georgkeller.bsky.social · Jul 30

i.e. if the coding space of your population can represent bimodal distributions (in a non-trivial way), so can the prediction errors. And regarding the role of prediction errors in driving plasticity, the problem is the same for gradient descent in backprop.

1 1

Georg Keller @georgkeller.bsky.social · Jul 30

Good point, but I don’t think that is different from other models of cortical function? Regarding the role of prediction errors in updating internal representations, the coding space of predictions, prediction errors, and internal representation just needs to match.

1 1

Georg Keller @georgkeller.bsky.social · Jul 29

Intuitively, I think this would make sense – imagine opening a cookie jar to find a mouse having eaten your last cookie, I suspect your brain will prioritize the unexpected presence of the mouse, to only later be irked by the unexpected absence of the cookie.

1

Georg Keller @georgkeller.bsky.social · Jul 29

My interpretation of this is that there likely is an inhibitory competition between negative and positive prediction errors that always ends up favoring positive prediction errors.

1 2

Georg Keller @georgkeller.bsky.social · Jul 29

What I think Sonja’s paper very nicely shows is that presenting both positive and negative prediction errors (i.e. a stimulus substitution), drives only a positive prediction error response.

1 2

Georg Keller @georgkeller.bsky.social · Jul 29

A negative prediction error neuron for stimulus surround (a la Rao&Ballard) might not respond to a visuomotor mismatch, etc.

1 2

Georg Keller @georgkeller.bsky.social · Jul 29

There could be dedicated prediction error neurons to the different predictions (e.g. a visuomotor prediction, an audio-visual prediction, a stimulus surround prediction). It is still unclear how different predictions are combined.

1

Georg Keller @georgkeller.bsky.social · Jul 29

Re 2) Also agreed. But there are multiple viable interpretations here. Assuming we let go of the idea of cortex as hierarchy (cortex is a hierarchy, like the earth is flat) – then there must are multiple predictive inputs.

1 2

Georg Keller @georgkeller.bsky.social · Jul 29

The notion of ‘generalized predictive coding’ sounds similar to the assumption that if something is predictable in principle, the brain must predict it. I don’t think there ever was good evidence to support that assumption?

2 2

Georg Keller @georgkeller.bsky.social · Jul 29

The two instances we have solid evidence for prediction errors are stimulus surround (i.e. statistics of natural images) and visuomotor coupling (i.e. physics of the world) – both likely behaviorally relevant.

1

Georg Keller @georgkeller.bsky.social · Jul 29

Re 1) Correct, and why would it? We can easily construct (artificial) examples of perfectly predictable sequences of images we would intuitively agree the brain likely does not predict (think e.g. sequences of random visual noise that repeats every X s or so).

1 1

Georg Keller @georgkeller.bsky.social · Jul 29

We had to learn (by trial and error) that even for low level audio-visual interactions, the mouse brain shows no signs of computing predictions or prediction errors unless the stimuli are also paired with a water reward or an air puff (pubmed.ncbi.nlm.nih.gov/34857950/).

A cortical circuit for audio-visual predictions - PubMed

Learned associations between stimuli in different sensory modalities can shape the way we perceive these stimuli. However, it is not well understood how these interactions are mediated or at what leve...

pubmed.ncbi.nlm.nih.gov

1 1 4

Georg Keller @georgkeller.bsky.social · Jul 29

And I suspect, if they were aversive, the brain might start predicting them – but they are probably just… boring? It is quite remarkable to see how quickly sensory responses to most stimuli ‘adapt’ away if the stimuli are presented repeatedly for days.

1 3

Georg Keller @georgkeller.bsky.social · Jul 29

I think that is the key question, and it can probably only be answered by doing the experiment (i.e. testing whether there are signals consistent with predictions and prediction errors). I suspect that without any relevance, passively presented experimental stimuli tend to be ignored by the brain.

1 2

Reposted by Georg Keller

eLife @elife.bsky.social · Jul 21

Ditching months-long delays for fast, constructive feedback.

This interview with @solygamagda.bsky.social dives into the experience of publishing with eLife and what it could mean for a more open and efficient future in science.

Publishing with eLife: “the future of science lies in greater transparency”

Neuroscientist Magdalena Solyga shares her latest study and her experience publishing with eLife.

buff.ly

9 15

Georg Keller @georgkeller.bsky.social · Jul 16

pubmed.ncbi.nlm.nih.gov/38424472/
pubmed.ncbi.nlm.nih.gov/30359606/
These are two I was involved in – sorry for the biased answer - there are probably better ones

Predictive Processing: A Circuit Approach to Psychosis - PubMed

Predictive processing is a computational framework that aims to explain how the brain processes sensory information by making predictions about the environment and minimizing prediction errors. It can...

pubmed.ncbi.nlm.nih.gov

1 3

Reposted by Georg Keller

Dan Goodman @neural-reckoning.org · Jul 16

Defending science in public we often talk about 'peer reviewed science'. But could this framing contribute to undermining trust in science and holding us back from improving the scientific process? Instead, let's talk about the work that has received the most thorough and transparent scrutiny? 🧪

2 8 22

Georg Keller @georgkeller.bsky.social · Jul 16

This they show the animal thousands of times and then test whether the pattern X G X G X G X (with a final X instead of the Y – a so called “global oddball”) elicits a prediction error response. They find it does not.

1