In a #ICML2026 position paper we argue a dynamical systems perspective is needed to drive time series models forward: arxiv.org/abs/2602.16864
For TS, we need to move away from transformers that do not respect a system’s dynamical structure, esp. if out-of-domain generalization & insight is sought.
Neural ODEs are great as continuous-time dynamical systems, but slow and tedious to train.
In a new #ICML2026 paper we intro a novel solver for continuous-time RNNs that does not rely on numerical integration. It's not only way faster & robust, but enables explicit analysis: arxiv.org/abs/2602.15649
DurstewitzLab
DurstewitzLab
Our new paper is out this week in Nature Neuroscience! www.nature.com/articles/s41... We built a BCI that works with the brain's natural geometry — and we found that people could learn to play a video game with their brains in <1 hr of training. This efficiency is groundbreaking & here's why:
Busch et al. use nonlinear neural manifolds to help humans gain rapid control over a noninvasive brain–computer interface, allowing them to learn how to play a video game with real-time fMRI neurofeed...
Some more Brain-Machine Interface shenanigans: this time we think we found evidence that an animal's agency in a task modulates hippocampal maps of that task.
www.biorxiv.org/content/10.6...
Even if they don't dismiss the evidence of its existence, a big chunk of neuroscientists don't seem to get a key question that representational drift raises.
The whole point is that responses reconfigure *without* loosing representational fidelity.
This means *of necessity* that any geometric/
