Arman Behrad

For Spanish researchers interested in postdoc positions: my lab at Amsterdam is looking for candidates to apply for a Ramon Areces postdoctoral fellowship, to work for 2 years on compneuro and digital brains. Deadline Feb 23. Please spread the word! More info: www.fundacionareces.es/fundacionare...

Built a domain-agnostic peak detection algorithm and now hunting for datasets with known/annotated peaks to test it on 👀
Any domain works—signals,bio,astro,finance, spectroscopy, etc.
Got data or know a benchmark? Would love pointers 🙌

#SignalProcessing #DataScience #TimeSeries #OpenData #Research

This book is a wonderful, synthetic and richly illustrated journey through the natural history of the vertebrate brain 🤩

A big thank you to the authors 🙏

"A major theme in the evolution of the telencephalon has been the emergence of novel pathways...

1/2

This work couldn’t have happened without my wonderful collaborators: @neurostrow.bsky.social and Ila Fiete (master minds behind the original DSA), @mmdtaha.bsky.social, Christian Beste, and @neuroprinciplist.bsky.social ; and support from the @cmc-lab.bsky.social .

There are also other fantastic tools around for comparing circuits/brains/models like the RSA developed by Nikolaus Kriegeskorte and many others (e.g., see the great work www.biorxiv.org/content/10.1... by @jbarbosa.org , and @itsneuronal.bsky.social )

Related: @neurostrow.bsky.social thread with @wtredman.bsky.social & Igor Mezic on extending dynamical-similarity ideas—an exciting direction for future DSA-style methods: bsky.app/profile/neur...

We also built 2 simple nonlinear systems (A, B) with identical eigenvalues but different eigenvectors. As expected, Wasserstein-based kwDSA struggles to separate them. All 3 fastDSA variants reliably distinguish A vs B (represented w/ MDS).

Under strong noise, we repeat the transformation tests (Plus kernelDMD+Wasserstein distance (kwDSA)). kwDSA highlights a key pitfall: relying mainly on eigenvalues (ignoring eigenvectors) can miss fine dynamical differences. fastDSA alternatives remain sensitive and perform well even at high noise.

Next we tested sensitivity to dynamical change by morphing a ring attractor into a line attractor (same model). fastDSA distances jump at the ring↔line transition, capturing topology change—unlike Procrustes—while being way faster than DSA.

We first tested whether fastDSA is invariant to purely geometric deformations—changes that preserve the same underlying dynamics and attractor topology. All 3 fastDSA variants are faithful to dynamics and remain stable across geometric deformations, while being computationally more efficient.

With different forms of noise, we showed how well the rank estimate supports DMD reconstruction. Across noise levels, the method detects the rank at the knee point automatically (with no tuning)

Our method efficiently estimates the rank of delay embeddings of a dynamical system. For example, on Lorenz trajectories projected to higher dimensions, the estimated order matches the true latent rank and aligns with AIC/BIC baselines.

We made DSA up to 150 times faster 🤯 by introducing 3 new optimization objectives and solvers to speed up the DSA alignment step. Instead of enforcing exact orthogonality at every iteration, we use faster formulations that approximate or penalize the constraint.

The original dynamic similarity analysis (DSA) developed by @neurostrow.bsky.social and Ila Fiete is a powerful method to compare trajectories of (nonlinear) neural dynamics between different datasets and models: arxiv.org/abs/2306.10168

Wanna compare dynamics across neural data, RNNs, or dynamical systems? We got a fast and furious method🏎️
The 1st preprint of my PhD 🥳 fast dynamical similarity analysis (fastDSA):
📜: arxiv.org/abs/2511.22828
💻: github.com/CMC-lab/fast...
I’ll be @cosynemeeting.bsky.social - happy to chat 😉

Way back in 1999, Kenji Doya sketched a big picture theory of the brain:

1️⃣The cerebellum is specialized for supervised learning
2️⃣The basal ganglia are for reinforcement learning
3️⃣The cerebral cortex is for unsupervised learning

How does this hold up in 2026? www.sciencedirect.com/science/arti...

From sensory to perceptual manifolds: The twist of neural geometry The brain uses geometric twists to expand neural dimensionality, thus untangling perception from sensation.

From sensory to perceptual manifolds: The twist of neural geometry
doi.org/10.1126/scia...
#neuroscience

Thank you for having me on BrainInspired, Paul @braininspired.bsky.social! It was such an honor to be on my favorite show—a rare place where we can leisurely talk about manifolds, latent circuits, power laws, and other esoteric ideas, and still be taken seriously in knowing they are all real.

🚨new work with the dream team @danakarca.bsky.social @loopyluppi.bsky.social @fatemehhadaeghi.bsky.social @stuartoldham.bsky.social @duncanastle.bsky.social
We use game theory and show the brain is not optimally wired for communication and there’s more to its story:
www.biorxiv.org/content/10.6...

Congratulations. 🎊

Our next paper on comparing dynamical systems (with special interest to artificial and biological neural networks) is out!! Joint work with @annhuang42.bsky.social , as well as @satpreetsingh.bsky.social , @leokoz8.bsky.social , Ila Fiete, and @kanakarajanphd.bsky.social : arxiv.org/pdf/2510.25943

If you are @ #BernsteinConference we have 4 posters, thread👇
P II 25, Tue 18–19:30: abstracts.g-node.org/conference/B...
P IV 5, Wed 14-15:30: abstracts.g-node.org/conference/B...
P II 25, Tue 18-19:30: abstracts.g-node.org/conference/B...
P I 11, Tue 16:30-18: abstracts.g-node.org/conference/B...

(4/4) How about a simulation of more naturalistic setting?

(3/4)
Next: biological networks.
We applied UNITE to V4 activity (PSTH) from a monkey doing a visual cognitive task. 🐒👁️

(2/4)
First test: an RNN solving a cognitive task.
UNITE identify state transitions without supervision. 🔍

How can we identify state transitions from neural data — in an unsupervised way? (1/4)
👉 Check out UNITE: Universal Neural-State Identification through Temporal Embeddings
📅 Oct 1 | #BernsteinConference
📍 Poster Session III, #14
abstracts.g-node.org/conference/B...

(4/4)
Finally, we highlight cases where FastDSA outperforms other dynamical similarity metrics—while remaining robust to noise.

(3/4)
Next, we present experiments showing why geometrical methods fail.
Both DSA and FastDSA achieve similar results—but FastDSA is 10–15× faster.

(2/4)
We first outline the general structure of the algorithm.

Curious about similarity metrics?
What if we could capture dynamical similarity that’s noise-robust and computationally efficient? (1/4)
Check out Fast Dynamical Similarity Analysis
📅 30 Sept | #BernsteinConference
📍 Poster Session I, #11

abstracts.g-node.org/conference/B...