Sign in

Jonathan Karin

@jonathankarin.bsky.social
7 followers 19 following 15 posts
PostsRepliesMedia
Jonathan Karin @jonathankarin.bsky.social · 08/12/2025
1/9 How do swarms (like fish schools) survive predator attacks? In our new paper at @PhysRevE, we try to break the "Detectability-Durability" trade-off with GNN-based generative modeling! with @zoepiran and @mornitzan.bsky.social journals.aps.org/pre/abstract...
journals.aps.org
120
Reposted by Jonathan Karin
arXiv q-bio.QM Quantitative Methods @qbioqm-bot.bsky.social · 08/07/2025
Jonathan Karin, Zoe Piran, Mor Nitzan: Enhancing Swarms Durability to Threats via Graph Signal Processing and GNN-based Generative Modeling arxiv.org/abs/2507.03039 arxiv.org/pdf/2507.03039 arxiv.org/html/2507.03039
023
Reposted by Jonathan Karin
Omer Karin @omerka.bsky.social · 20/12/2024
Happy to share this- we analyze the hierarchical organization of attractor landscapes in dense associative memories and how it can facilitate transitions between attractor states. We use this to explain the control of transitions between stem and differentiated identities. arxiv.org/abs/2412.11336
arxiv.org
Hierarchical Control of State Transitions in Dense Associative Memories
We analyze a generalization of Modern Hopfield networks that emerges naturally in the gene regulatory networks that control cellular identity. Here gene expression patterns correspond to stored memori...
031
Jonathan Karin @jonathankarin.bsky.social · 08/12/2024
1/3 Annotatability is out in Nature Computational Science! With Reshef Mintz, @ravehlab.bsky.social and @mornitzan.bsky.social www.nature.com/articles/s43...
nature.com
Interpreting single-cell and spatial omics data using deep neural network training dynamics - Nature Computational Science
The Annotatability framework analyzes neural network training dynamics to interpret single-cell and spatial omics data. It identifies erroneous annotations and ambiguous cell states, infers trajectori...
121
Jonathan Karin @jonathankarin.bsky.social · 05/12/2024
1/3 Annotatability is out in Nature Computational Science! With Reshef Mintz, @ravehlab.bsky.social and @mor_nitzan www.nature.com/articles/s43... .
nature.com
Interpreting single-cell and spatial omics data using deep neural network training dynamics - Nature Computational Science
The Annotatability framework analyzes neural network training dynamics to interpret single-cell and spatial omics data. It identifies erroneous annotations and ambiguous cell states, infers trajectori...
121