Shahab Bakhtiari

shahabbakht.bsky.social

06 January 2026 21:26

Is it just my timeline, or are we seeing a spike in loose neuro-based AI ideation again? I’m seeing way more references to specific mechanisms by AI folks (hippocampus, oscillations, etc.). It's often handwavy, but interesting to see the pendulum swing back after years of fading interest. 🧠🤖

Shahab Bakhtiari

shahabbakht.bsky.social

04 September 2025 19:58

Unrelated note: The IBL paper isn't an evidence against specialization.

Timothy O'Leary

timothyoleary.bsky.social

04 September 2025 16:43

Finding correlates of the same signal in different brain areas is not evidence against specialisation, it is evidence for interconnection. When I flush my toilet, the level in the tank and the flow rate in the supply pipe become perfectly correlated. Yet these toilet areas have distinct functions.

Shahab Bakhtiari

shahabbakht.bsky.social

14 May 2025 20:15

Check out our new paper! Vision models often struggle with learning both transformation-invariant and -equivariant representations at the same time. @hafezghm.bsky.social shows that self-supervised prediction with proper inductive biases achieves both simultaneously. (1/4) #MLSky #NeuroAI

Beth Lucas

ewhizabeth.bsky.social

02 May 2025 14:31

We have been conducting calcium imaging in Rspo2+ neurons in the amygdala, which are reported to encode negative valence. I am totally obsessed with the Rspo2-Cre mouse line! Look at the stream of axons leaving the amygdala to their final destinations! #FluorescenceFriday #Neuroskyence