This is potentially relevant for understanding impaired reading (e.g., in dyslexic readers) as behaviors of early learning in beginning readers are often conserved. Consequently, we argue that understanding these is highly important.
We concluded that neuro-cognitive mechanistic models are useful not only for modeling behavioral response patterns across species but also for describing the internal representations implemented, especially during initial learning periods.
The SLR is transparent, allowing us to investigate the representations used (i.e., computational phenotypes), showing that visual representations are informative during early learning. However, both humans and baboons rely on larger orthographic representation units as script familiarity increases.
We then examined how well each model simulated response patterns and found that the much simpler SLR model had the highest prediction overlap for both humans and baboons.
Using the two datasets, we applied a domain-specific neuro-cognitive model which follows the principles of predictive coding, the Speechless Reader (SLR), and two connectionist models, CORnet-Z and ResNet-18. All model simulations were exposed to the exact same stimuli as humans and baboons.
In the study, we simulated how humans and baboons learn to make orthographic decisions. We used a dataset from Eisenhauer et al. (2019) where humans learned a set of N = 60 pseudowords and Grainger et al. (2012) where baboons learned to recognize english words.
Happy to share that my recent paper with @benjamingagl.bsky.social "Understanding the Neuro-Cognitive Mechanisms of Orthographic Learning in Humans and Baboons" is now published in Springer Nature: Computational Brain and Behavior (link.springer.com/article/10.1...)
We're excited to announce today's #GeWoNN Journal Club Meeting! Our guest speaker, Daria Kostanian, will talk about "Development of coarse- and orientation-tuning for native letters in children" (Paper link: tinyurl.com/28fb4z89)
TODAY!! 24-3-2026, 11-12 am (CET)
Zoom link👇
tinyurl.com/3m9rfuyx
Happy New Year! 🥳 We're excited to announce January's #GeWoNN Journal Club Meeting! Our guest speaker, Dr Kurt Winsler, will talk about "Experience-Dependent Changes in Visual Processing of Letters".
Date: 21-1-2026, 9-10 am (CET)
Register here for the Zoom link 👇
forms.gle/e25X4FoQ2VUh...
Will this event be livestreamed, or will a recording of the debate be uploaded? Sounds super interesting and I would be interested to see it.
Sign up for the next virtual #GeWoNN Journal Club Meeting!
We are thrilled to have @juliafstrand.bsky.social give us a guest talk on error prevention in psychological research!
Linguists, psychologists, and language researchers are all welcome! 😊
Register here: forms.gle/YoZpPxrRxXcG...
We fitted all computational models on the learning data from humans & baboons to see which best simulates orthographic learning🧠🐒. The results show, that the mechanistic SLR aligns closest with the data. Be sure to come by on wednesday if you’re curious - Im happy to chat and discuss our study!
I will present my poster on the neuro-cognitive mechanisms of orthographic learning in humans 🚶♀️ and baboons 🐒 on wednesday at #CCN (Poster B34). In our study, we compared CORnet-Z and ResNet 18, with a transparent, mechanistic model of reading: The Speechless Reader Model (SLR).
Understanding the Neuro-Cognitive Mechanisms of Orthographic Learning in Humans and Baboons: A Comparative Study Using Domain-Specific Mechanistic and Domain-General Connectionist Models www.biorxiv.org/content/10.1101/2025.05....
Welcome to join our journal club discussion tomorrow! 😄
Join our next journal club meeting on May 22 (Thur), 11am-12pm (CEST)! We'll discuss: @marcelbinz.bsky.social et al.'s (2024) "Centaur: a foundation model of human cognition"!
😊 Moderator: @ksenijamisic.bsky.social
👇 See link for more info. DM or email us for the Zoom link!
Latest updates on the neural basis of invariant letter position coding in the brain
Three more days for registration! 😊
🚨 TRUST Open Lecture – Join Us! 🚨
Is #standardisation in research boosting #replicability or slowing scientific progress? 🤔
📅 March 31 (Mon), 2025
⏰ 16:00–18:00 CET
📍 Hörsaal H112, IBW-Gebäude, University of Cologne
💻 Live-stream available! Registration info here:
sites.google.com/view/gewonn/...
We're collecting opinions on #standardisation of #research processes before an event next month. Please feel free to vote & share: www.linkedin.com/posts/leungy...
The public event will be a #panel #discussion & will be live streamed. If you're interested in learning more, please get in touch!
Just left the #Teap2025 conference with a contribution to the „From retina to word form“ symposium, which I was lucky to co-organize with @jacketaylor.bsky.social.
Super interesting contributions by all attendees!
@benjamingagl.bsky.social , @Aakash Agrawal, @merferlo.bsky.social
Check out our symposium on levels of #OrthographicProcessing at #teap2025!
March 12 | 11:00–12:30
Chairs: @jacketaylor.bsky.social, @jnpauli.bsky.social
Speakers: Aakash Agrawal, @merferlo.bsky.social, Jack E. Taylor, Janos Pauli, @benjamingagl.bsky.social
Congratulations to our #GeWoNN members on their accepted symposia at the #teap2025 Conference! Special thanks to the organising committee @teap2025.bsky.social, and we're looking forward to the event next week in Frankfurt!
Check out our website below for more news on our symposia!
Hey everyone, we have organized a lecture and panel discussion on the topic "Standardization in Research: Increasing Replicability or Stalling Scientific Progress?" in Cologne on March 31st: sites.google.com/view/gewonn/... @gewonnofficial.bsky.social @unicologne.bsky.social
It's all gone. They've taken down all the CDC datasets from data.cdc.gov
Data that's critical for public health & research. Data paid for by citizens. That the government now censors.
"Do not fear those who tell the truth. Fear those who try to hide it. They are willing to sacrifice other things."
Hey Bsky friends on #neuroskyence! Very excited to share our
@iclr-conf.bsky.social paper: TopoNets! High-performing vision and language models with brain-like topography! Expertly led by grad student Mayukh and Mainak! A brief thread...
Researchers paid nearly $9 billion to publishers between 2019 and 2023 so you could read our work. Why? Why??
The FDA has scrubbed from its website guidance on conducting medical trials with representative numbers of men/women and racial/ethnic minorities. these aren't symbolic issues. Drugs affect men/men, people with different genetic backgrounds differently. talkingpointsmemo.com/edblog/medic...
Forderungen von #IchBinHanna an neue Bundesregierung:
1️⃣ Postdocs raus aus #WissZeitVG
2️⃣ #Befristungshöchstquote
3️⃣ Flexible Personalkategorien neben Profs
4️⃣ Departmentstrukturen
5️⃣ mind. 4 J. Vertragslaufzeit für Promotion
6️⃣ Exzellenz- & Großteil d. DFG-Mittel in Grundfinanzierung
#HannaWählt2025
