Dirk Gütlin
@gutlin.bsky.social
I post mainly about Neuroscience, Machine Learning, Complex Systems, or Stats papers.
Working on neural learning /w @auksz.bsky.social CCNB/BCCN/Free University Berlin.
I also play bass in a pop punk band:
https://linktr.ee/goodviewsbadnews
Working on neural learning /w @auksz.bsky.social CCNB/BCCN/Free University Berlin.
I also play bass in a pop punk band:
https://linktr.ee/goodviewsbadnews
July 8, 2025 at 6:55 PM
A helpful rule:
If you are punching down, you probably are on the morally wrong side.
This is because in a conflict, the group in power in almost all cases holds the cards to decide whether to solve the conflict by ethical or unethical means.
It's impressive on how many levels this rule applies.
If you are punching down, you probably are on the morally wrong side.
This is because in a conflict, the group in power in almost all cases holds the cards to decide whether to solve the conflict by ethical or unethical means.
It's impressive on how many levels this rule applies.
July 6, 2025 at 7:33 AM
A helpful rule:
If you are punching down, you probably are on the morally wrong side.
This is because in a conflict, the group in power in almost all cases holds the cards to decide whether to solve the conflict by ethical or unethical means.
It's impressive on how many levels this rule applies.
If you are punching down, you probably are on the morally wrong side.
This is because in a conflict, the group in power in almost all cases holds the cards to decide whether to solve the conflict by ethical or unethical means.
It's impressive on how many levels this rule applies.
I'm a German. I usually shut my mouth on this. But right now, I can't.
This is not a war, this is a massacre.
And it's heartbreaking. We will have to take responsibility for this. If we learnt anything from history then it's to not stay silent now.
(Stats from Wikipedia: "Gaza War")
This is not a war, this is a massacre.
And it's heartbreaking. We will have to take responsibility for this. If we learnt anything from history then it's to not stay silent now.
(Stats from Wikipedia: "Gaza War")
May 21, 2025 at 9:21 AM
I'm a German. I usually shut my mouth on this. But right now, I can't.
This is not a war, this is a massacre.
And it's heartbreaking. We will have to take responsibility for this. If we learnt anything from history then it's to not stay silent now.
(Stats from Wikipedia: "Gaza War")
This is not a war, this is a massacre.
And it's heartbreaking. We will have to take responsibility for this. If we learnt anything from history then it's to not stay silent now.
(Stats from Wikipedia: "Gaza War")
If you keep using #Twitter, be aware that the content pushed to your feed is extremely distorted towards right-wing/anti-immigrant ideology. This paper provides clear evidence for that:
"Political Biases on X before the 2025 German Federal Election"
arxiv.org/abs/2503.02888
"Political Biases on X before the 2025 German Federal Election"
arxiv.org/abs/2503.02888
March 7, 2025 at 3:32 PM
If you keep using #Twitter, be aware that the content pushed to your feed is extremely distorted towards right-wing/anti-immigrant ideology. This paper provides clear evidence for that:
"Political Biases on X before the 2025 German Federal Election"
arxiv.org/abs/2503.02888
"Political Biases on X before the 2025 German Federal Election"
arxiv.org/abs/2503.02888
#Preprint alert!
🧠📈 🧠🤖
Ever wondered how well Predictive Coding can be implemented in Artificial Neural Nets? Do they really learn as the theory predicts?
In our new preprint with @auksz.bsky.social, we test two PC algorithms & find that not all behave as expected!
www.biorxiv.org/content/10.1...
🧠📈 🧠🤖
Ever wondered how well Predictive Coding can be implemented in Artificial Neural Nets? Do they really learn as the theory predicts?
In our new preprint with @auksz.bsky.social, we test two PC algorithms & find that not all behave as expected!
www.biorxiv.org/content/10.1...
February 6, 2025 at 2:55 PM
#Preprint alert!
🧠📈 🧠🤖
Ever wondered how well Predictive Coding can be implemented in Artificial Neural Nets? Do they really learn as the theory predicts?
In our new preprint with @auksz.bsky.social, we test two PC algorithms & find that not all behave as expected!
www.biorxiv.org/content/10.1...
🧠📈 🧠🤖
Ever wondered how well Predictive Coding can be implemented in Artificial Neural Nets? Do they really learn as the theory predicts?
In our new preprint with @auksz.bsky.social, we test two PC algorithms & find that not all behave as expected!
www.biorxiv.org/content/10.1...
A universal hippocampal memory code across animals and environments
#neuroscience #compneuro
www.biorxiv.org/content/10.1...
#neuroscience #compneuro
www.biorxiv.org/content/10.1...
November 29, 2024 at 2:25 PM
A universal hippocampal memory code across animals and environments
#neuroscience #compneuro
www.biorxiv.org/content/10.1...
#neuroscience #compneuro
www.biorxiv.org/content/10.1...
B-cosification: Transforming Deep Neural Networks to be Inherently Interpretable
Cool way of retraining trained ANN to make them more interpretable
#MLSky #stats
arxiv.org/abs/2411.00715
Cool way of retraining trained ANN to make them more interpretable
#MLSky #stats
arxiv.org/abs/2411.00715
November 4, 2024 at 10:42 AM
B-cosification: Transforming Deep Neural Networks to be Inherently Interpretable
Cool way of retraining trained ANN to make them more interpretable
#MLSky #stats
arxiv.org/abs/2411.00715
Cool way of retraining trained ANN to make them more interpretable
#MLSky #stats
arxiv.org/abs/2411.00715
Equivalence between representational similarity analysis, centered kernel alignment, and canonical correlations analysis
#stats #compneuro #neuroscience
www.biorxiv.org/content/10.1...
#stats #compneuro #neuroscience
www.biorxiv.org/content/10.1...
November 1, 2024 at 11:34 AM
Equivalence between representational similarity analysis, centered kernel alignment, and canonical correlations analysis
#stats #compneuro #neuroscience
www.biorxiv.org/content/10.1...
#stats #compneuro #neuroscience
www.biorxiv.org/content/10.1...
March 2, 2024 at 2:14 PM