Melanie Tschiersch
banner
mtschiersch.bsky.social
Melanie Tschiersch
@mtschiersch.bsky.social
Computational Neuroscience PhD student in Barcelona. Working on Data Analysis of behavior and neural recordings and computational modeling of interhemispheric working memory
Thank you so much, Seán! It's been a long time in the making but finally getting somewhere!
January 21, 2025 at 11:05 AM
🗯️Taken together, this shows that prefrontal hemispheres can adjust to varying task demands with a simple underlying architecture!

If you're interested in more details, check out the preprint or send me a message!

Also a HUGE shout-out to everyone involved.It's been (and will continue) fun!🎉 END/🧵
January 20, 2025 at 12:10 PM
Well... turns out REDUNDANT hemispheres can also replicate this effect!! So it seems that redundant, weakly connected hemispheres can provide increased precision (redundancy) when WM demands are low and increased capacity (weak connections) when WM demands are high. 7/🧵
a man is making a funny face while holding his finger to his head .
ALT: a man is making a funny face while holding his finger to his head .
media.tenor.com
January 20, 2025 at 12:10 PM
When inputs were slightly lateralized, the model showed that each hemisphere stores the stronger contra- over a weaker ipsilateral input when WM capacity is reached. Remember from 2/ where I said the bilateral field advantage was seen as evidence for non-redundant, independent hemispheres?? 6/🧵
January 20, 2025 at 12:10 PM
This got us thinking: This model can represent the same memory redundantly, but also maintains history drift privately, allowing for private and shared memories. What does this mean when multiple items need to be remembered?🤔 5/🧵
January 20, 2025 at 12:10 PM
we modeled two redundant spiking bump attractor networks with inter-area connections and replicated the data if connections were tuned and weak. Furthermore, 🖥️model + 🐒data were in agreement that weak connections lead to private history attraction (serial dependence) in each hemisphere. 4/🧵
January 20, 2025 at 12:10 PM
We first found that both hemispheres stored (bump-like FR in the figure) and guided behavior for the full visual field suggesting redundant hemispheres. Additionally, neural traces were only weakly correlated across hemispheres suggesting weak coordination between hemispheres... 3/🧵
January 20, 2025 at 12:10 PM
How do prefrontal hemispheres support continuous spatial WM when visual inputs come strongly lateralized?🤔
Previous studies suggest:
1️⃣ Independent hemispheres:Larger capacity across hemifields (bilateral field advantage)?
2️⃣ Redundant hemispheres:Unilateral opto-disruption doesn’t impair behavior?2/🧵
January 20, 2025 at 12:10 PM
good luck! Tagesschau says they are sceptical it will even arrive at all... but fingers crossed!
January 20, 2025 at 9:57 AM