Don’t miss these exciting perspectives at the intersection of AI and neuroscience!
📅 8th September
🔗 Zoom link: ethz.zoom.us/j/66426188160
4/4

🎙️ Austin Meek
📢 “Inducing Human-like Biases in Moral Reasoning Language Models”
Research scholar in the ML Alignment & Theory Scholars program and PhD student at the University of Delaware, studying chain-of-thought faithfulness, brain–model alignment and interpretability.
3/4

🎙️ Loek van Rossem
📢 “When Representations Align: Universality in Representation Learning Dynamics”
Loek is a PhD researcher at the Gatsby Unit, UCL. His work focuses on deep learning theory, universal representation learning, and mechanistic interpretability
2/4

👩‍🔬 About the speaker:
 Dr. @irisgroen.bsky.social is Assistant Professor at the Video & Image Sense lab (UvA, Amsterdam). Her work bridges AI & cognitive neuroscience, studying how the brain turns light into meaning.
🌐 www.irisgroen.com
4/4

Dr. Groen’s talk explores how we can push beyond object recognition:

🔹 Scene affordance perception

🔹 Video perception

🔹 Data-driven alignment via brain-guided image generation.
A new step toward AI that mirrors real-world human perception.
3/4

For years, deep neural networks (DNNs) amazed scientists by predicting neural responses in the visual cortex. 👀
But most studies only tested this on static, isolated objects.
Real life? It's way more complex. 🌆🎥
2/4

For years, deep neural networks (DNNs) amazed scientists by predicting neural responses in the visual cortex. 👀
But most studies only tested this on static, isolated objects.
Real life? It's way more complex. 🌆🎥
2/4