Mathematical Framework for Neural Population Coding Proposed in arXiv Paper
A new study on arXiv (ID: 2309.02332) offers an innovative mathematical approach to better understand how information is processed in the mammalian brain. The research develops an algebra involving convex cones based on a model of an individual plastic neuron, which helps to describe activity across neuron populations. By viewing these populations as active elements in this algebraic framework, researchers can create concise algebraic formulas. These formulas support functions like specialization, generalization, novelty detection, reducing dimensions, inverse modeling, prediction, and associative memory. The authors aim to bridge the gap between the behavior of single neurons and the computations at the population level, providing a theoretical basis for future studies.
Key facts
- Paper ID: arXiv:2309.02332
- Announcement type: replace-cross
- Abstract introduces a mathematical framework for neural population coding
- Framework uses an algebra of convex cones
- Derived from a mechanistic model of a single plastic neuron
- Interconnected populations realize compact algebraic expressions
- Functions include specialization, generalization, novelty detection, dimensionality reduction, inverse modeling, prediction, and associative memory
- Published on arXiv
Entities
Institutions
- arXiv