Neural spiking for causal inference and learning.
When a neuron is driven beyond its threshold, it spikes. The fact that it does not communicate its continuous membrane potential is usually seen as a computational liability. Here we show that this spiking mechanism allows neurons to produce an unbiased estimate of their causal influence, and a way...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2023-04-01
|
| Series: | PLoS Computational Biology |
| Online Access: | https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1011005&type=printable |
| _version_ | 1826864324398809088 |
|---|---|
| author | Benjamin James Lansdell Konrad Paul Kording |
| author_facet | Benjamin James Lansdell Konrad Paul Kording |
| author_sort | Benjamin James Lansdell |
| collection | DOAJ |
| description | When a neuron is driven beyond its threshold, it spikes. The fact that it does not communicate its continuous membrane potential is usually seen as a computational liability. Here we show that this spiking mechanism allows neurons to produce an unbiased estimate of their causal influence, and a way of approximating gradient descent-based learning. Importantly, neither activity of upstream neurons, which act as confounders, nor downstream non-linearities bias the results. We show how spiking enables neurons to solve causal estimation problems and that local plasticity can approximate gradient descent using spike discontinuity learning. |
| first_indexed | 2024-04-09T13:55:27Z |
| format | Article |
| id | doaj.art-1ad52e71fa974d33a6e506f794398dcc |
| institution | Directory Open Access Journal |
| issn | 1553-734X 1553-7358 |
| language | English |
| last_indexed | 2025-02-16T21:19:53Z |
| publishDate | 2023-04-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS Computational Biology |
| spelling | doaj.art-1ad52e71fa974d33a6e506f794398dcc2025-01-18T05:30:48ZengPublic Library of Science (PLoS)PLoS Computational Biology1553-734X1553-73582023-04-01194e101100510.1371/journal.pcbi.1011005Neural spiking for causal inference and learning.Benjamin James LansdellKonrad Paul KordingWhen a neuron is driven beyond its threshold, it spikes. The fact that it does not communicate its continuous membrane potential is usually seen as a computational liability. Here we show that this spiking mechanism allows neurons to produce an unbiased estimate of their causal influence, and a way of approximating gradient descent-based learning. Importantly, neither activity of upstream neurons, which act as confounders, nor downstream non-linearities bias the results. We show how spiking enables neurons to solve causal estimation problems and that local plasticity can approximate gradient descent using spike discontinuity learning.https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1011005&type=printable |
| spellingShingle | Benjamin James Lansdell Konrad Paul Kording Neural spiking for causal inference and learning. PLoS Computational Biology |
| title | Neural spiking for causal inference and learning. |
| title_full | Neural spiking for causal inference and learning. |
| title_fullStr | Neural spiking for causal inference and learning. |
| title_full_unstemmed | Neural spiking for causal inference and learning. |
| title_short | Neural spiking for causal inference and learning. |
| title_sort | neural spiking for causal inference and learning |
| url | https://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1011005&type=printable |
| work_keys_str_mv | AT benjaminjameslansdell neuralspikingforcausalinferenceandlearning AT konradpaulkording neuralspikingforcausalinferenceandlearning |