Computational neuroscientist in the connectionist tradition.
https://ari-benjamin.com
Ari Benjamin
Loading...
More of this type of content please!
Ari Benjamin
Nice summary of our recent work on neural population geometry and generalization across multiple tasks with shared latent structure.
Thank you @natmesanash.bsky.social and @thetransmitter.bsky.social for the great article!
Paper: www.nature.com/articles/s41...
Interesting – this is what you need for synaptic plasticity to solve "within-neuron" credit assignment. Backpropagating APs, elevating Ca2+ above the plasticity threshold, should preferentially enter dendrites which recently activated.
Is this the first evidence of this?
Creating scientists is half the point — as is inspiring scientific thinking in the rest of society, 'beyond the tower'.
If this is why we can't be automated, maybe the next 10 years can be a wake-up call to reach beyond the academy a bit more.
Talk of automating science reminds me there are two answers to "what is science for?"
If it's all 'solutions to problems', then automate away. There's no tear to shed.
But science is also: the anchor of secular society; our way of finding meaning in an infinite and delightful universe
This tool is extremely helpful and I think all scientists could benefit from it. Good at all stages of the research process, especially the very beginning.
It's also a great of example of how we *should* be building AI tools for science. Not to replace scientific thinking, but to hone it.
It is possible to work on neural networks as a neuroscientist and actively oppose how AI is developing and how it will affect society.
What do linear RNNs learn first? I made a highly-interactive blog post to explore this question, since it's more fun to play around with widgets than to read a pdf. ari-benjamin.com/rnn-modes/
Great way to remind yourself what a transfer function is!
10 reasons to be discouraged from a career in academia in one post
Especially relevant as a counterexample to OpenAI’s Prism tool, which will only worsen the science slop problem.
The point isn’t to be anti-AI. It’s to demand a better philosophy behind the tools we build
