Discussion
Loading...

#Tag

Log in
  • About
  • Code of conduct
  • Privacy
  • Users
  • Instances
  • About Bonfire
Fabrizio Musacchio
Fabrizio Musacchio
@FabMusacchio@mastodon.social  路  activity timestamp 2 weeks ago

馃 New paper by Ishida et al who show how #neurons in the #Drosophila central complex implement vector inversion via #calcium spikes.

A single #NeuronalPopulation can flip the sign of its encoded vector by switching biophysical #spiking modes, enabling coordinate transformations through #PopulationDynamics rather than circuit switching.

Cool as it shows that computation is not imposed by the circuit, but emerges from the neuron鈥檚 own dynamics.

馃實https://doi.org/10.1016/j.cell.2025.11.040

#Neuroscience #CompNeuro

Diagram showing how single neuron activity influences population activity, with vectors, signals, and neural pathways. Graphical abstract of the paper.
Diagram showing how single neuron activity influences population activity, with vectors, signals, and neural pathways. Graphical abstract of the paper.
Diagram showing how single neuron activity influences population activity, with vectors, signals, and neural pathways. Graphical abstract of the paper.

Neuronal calcium spikes enable vector inversion in the Drosophila brain

  • Copy link
  • Flag this post
  • Block
Fabrizio Musacchio
Fabrizio Musacchio
@pixeltracker@sigmoid.social  路  activity timestamp 2 months ago

馃 New preprint by Tilbury et al: Characterizing #NeuronalPopulation geometry with #AI equation discovery

The approach generates & evaluates 100s of candidate equations, finding "peaky" non-Gaussian tuning functions whose Fourier structure matches power-law dimensionality observed in real #V1 pops. Links shape of single- #neuron tuning to #PopulationLevel geometry using both data fits & analytical derivations.

馃實 https://doi.org/10.1101/2025.11.12.688086

#CompNeuro #Neuroscience #NeuralCoding #PopulationDynamics

2 media
Fig. 1. Oriented stimuli produce a high-dimensional population code not captured by standard tuning curve models.
Fig. 1. Oriented stimuli produce a high-dimensional population code not captured by standard tuning curve models.
Fig. 1. Oriented stimuli produce a high-dimensional population code not captured by standard tuning curve models.
Fig. 5. Effect of tuning peakiness on a simulated hyperacuity task.
Fig. 5. Effect of tuning peakiness on a simulated hyperacuity task.
Fig. 5. Effect of tuning peakiness on a simulated hyperacuity task.

Characterizing neuronal population geometry with AI equation discovery

  • Copy link
  • Flag this post
  • Block

bonfire.cafe

A space for Bonfire maintainers and contributors to communicate

bonfire.cafe: About 路 Code of conduct 路 Privacy 路 Users 路 Instances
Bonfire social 路 1.0.1 no JS en
Automatic federation enabled
Log in
  • Explore
  • About
  • Members
  • Code of Conduct