Excited to share that our paper got accepted @NeurIPS! 🎉

We introduce TRACE, a new contrastive learning method for visualizing multi-trial neural time-series data.

📊 2D visualizations revealing biological patterns
🔍 Automatic artifact detection
🔬 Works for calcium imaging & electrophysiology data

We found that:
🔹The biophys model can classify digits across different contrast & luminance levels
🔹Models w photoreceptor adaptation + bipolar cell ion channels are not improved by adding horizontal cells
🔹The biophys model maintains performance at contrasts outside of its training distribution

We built a highly-detailed, mechanistic model of cells in the mouse outer retina. We then constrained our model parameters with experimental data and a task (mnist classification at different contrast levels)…

Excited to present at #NeurIPS this week on my work building biophysical models of the outer retina! If you’re around come check it out or just say hi, poster #2015, Friday 11-2🎉

You can find the full paper here: openreview.net/forum?id=ayj...

In brief🧵:

Jaxley: differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics - Nature Methods Jaxley is a versatile platform for biophysical modeling in neuroscience. It allows efficiently simulating large-scale biophysical models on CPUs, GPUs and TPUs. Model parameters can be optimized with ...

Our work on training biophysical models with Jaxley is now out in @natmethods.nature.com. Led by @deismic.bsky.social, with @philipp.hertie.ai, @ppjgoncalves.bsky.social & @jakhmack.bsky.social et al.

Paper: www.nature.com/articles/s41...

At the European Retina Meeting and interested in biophysical modeling of the retina? Come stop by my poster (#081) tomorrow!

Agreed!

So excited for this tutorial! There's still time to submit abstracts for #CNS2025 in Florence😍