Human visual cortex representations may be much higher-dimensional than earlier work suggested, but are these higher dimensions of cortical activity actually relevant to behavior? Our new paper tackles this by studying how different people experience the same movies. 🧵 www.cell.com/current-biol...
High dimensional representations in the visual cortex, new paper from our lab, check it out!
Dimensionality reduction may be the wrong approach to understanding neural representations. Our new paper shows that across human visual cortex, dimensionality is unbounded and scales with dataset size—we show this across nearly four orders of magnitude. journals.plos.org/ploscompbiol...
Our modeling framework would offer a new avenue for understanding the computational principles of synaptic plasticity and learning in the brain. Research at HHMI Janelia, with fantastic collaborators Danil Tyulmankov, Adithya Rajagopalan, Glenn Turner, James Fitzgerald and @janfunkey.bsky.social!
We applied our technique to behavioral data from Drosophila in a probabilistic reward-learning experiment. Our findings reveal an active forgetting component in reward learning in flies 🪰, improving predictive accuracy over previous models. (4/5)
This method uncovers complex rules inducing long nonlinear time dependencies, involving factors like postsynaptic activity and current synaptic weights. We validate it through simulations, successfully recovering known rules like Oja’s and more intricate ones. (3/5)
website: yashsmehta.com/plasticity-p... Our approach approximates plasticity rules using parameterized functions—either truncated Taylor series for theoretical insights or multilayer perceptrons. We optimize these parameters via gradient descent over entire trajectories to match observed data (2/5)
🚀 Excited to share our paper has been accepted at #NeurIPS! 🎉 We developed a deep learning framework that infers local learning algorithms in the brain by fitting behavioral or neural activity trajectories during learning. We validate on synthetic data and tested on 🪰 behavioral data (1/5 🧵)
Interesting approach to estimate the physiological update rules of synapses: yashsmehta.com/plasticity-p...
🙌🏼🙌🏼
Thank you, Konrad!
