Shivashriganesh Mahato (@shivashriganesh)

Shivashriganesh Mahato @shivashriganesh Sep 21

We’re excited to share BrainWideBench, a new benchmark for evaluating how models of neural activity generalize across animals and tasks!

A major bottleneck in modeling neural data today is rigorous and standardized evaluation. Modelling techniques need to be tested under multiple regimes and downstream tasks.

📄 Check out the paper here! arxiv.org/abs/2609.22064

Special shoutout to @alx_adr, who co-led this work with me and contributed equally as co-first authors!

0:34

Shivashriganesh Mahato @shivashriganesh Dec 2, 2025

Check out this awesome new #NeurIPS2025 paper! NuCLR learns neuron-level representations that transfer zero-shot, a key step towards models that generalize across subjects without finetuning. Excited to see how this can extend to pop-level recordings and neuro-foundation models!

Vinam Arora Verified account @vinam_arora Dec 2, 2025

Excited to share our #NeurIPS2025 work: NuCLR, a framework for learning neuron-level representations 🧠 These embeddings capture the biological identity of neurons and work out-of-the-box on new animals; no finetuning needed 💃 This offers some of the first evidence that large-scale neuroscience models can truly generalize across animals.

Paper: arxiv.org/abs/2512.01199
Code: github.com/nerdslab/nuclr

If you are at NeurIPS in San Diego, come find us at Poster Session 5 (11am-3pm PT, Exhibit Hall C,D,E, # 2107) 🎉

1/x 🧵

View quoted post on X

Shivashriganesh Mahato

Latest posts 2