Inferring neural activity before plasticity as a foundation for learning beyond backpropagation
For both humans and machines, the essence of learning is to pinpoint which components in its information processing pipeline are responsible for an error in its output, a challenge that is known as ‘credit assignment’. It has long been assumed that credit assignment is best solved by backpropagation...
| Main Authors: | , , , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Springer Nature
2024
|
| _version_ | 1811140358837370880 |
|---|---|
| author | Song, Y Millidge, B Salvatori, T Lukasiewicz, T Xu, Z Bogacz, R |
| author_facet | Song, Y Millidge, B Salvatori, T Lukasiewicz, T Xu, Z Bogacz, R |
| author_sort | Song, Y |
| collection | OXFORD |
| description | For both humans and machines, the essence of learning is to pinpoint which components in its information processing pipeline are responsible for an error in its output, a challenge that is known as ‘credit assignment’. It has long been assumed that credit assignment is best solved by backpropagation, which is also the foundation of modern machine learning. Here, we set out a fundamentally different principle on credit assignment called ‘prospective configuration’. In prospective configuration, the network first infers the pattern of neural activity that should result from learning, and then the synaptic weights are modified to consolidate the change in neural activity. We demonstrate that this distinct mechanism, in contrast to backpropagation, (1) underlies learning in a well-established family of models of cortical circuits, (2) enables learning that is more efficient and effective in many contexts faced by biological organisms and (3) reproduces surprising patterns of neural activity and behavior observed in diverse human and rat learning experiments. |
| first_indexed | 2024-03-07T08:23:25Z |
| format | Journal article |
| id | oxford-uuid:8de90c12-c6c9-4b99-a7c2-2c5f9d410371 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-09-25T04:20:43Z |
| publishDate | 2024 |
| publisher | Springer Nature |
| record_format | dspace |
| spelling | oxford-uuid:8de90c12-c6c9-4b99-a7c2-2c5f9d4103712024-08-07T09:45:58ZInferring neural activity before plasticity as a foundation for learning beyond backpropagationJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:8de90c12-c6c9-4b99-a7c2-2c5f9d410371EnglishSymplectic ElementsSpringer Nature2024Song, YMillidge, BSalvatori, TLukasiewicz, TXu, ZBogacz, RFor both humans and machines, the essence of learning is to pinpoint which components in its information processing pipeline are responsible for an error in its output, a challenge that is known as ‘credit assignment’. It has long been assumed that credit assignment is best solved by backpropagation, which is also the foundation of modern machine learning. Here, we set out a fundamentally different principle on credit assignment called ‘prospective configuration’. In prospective configuration, the network first infers the pattern of neural activity that should result from learning, and then the synaptic weights are modified to consolidate the change in neural activity. We demonstrate that this distinct mechanism, in contrast to backpropagation, (1) underlies learning in a well-established family of models of cortical circuits, (2) enables learning that is more efficient and effective in many contexts faced by biological organisms and (3) reproduces surprising patterns of neural activity and behavior observed in diverse human and rat learning experiments. |
| spellingShingle | Song, Y Millidge, B Salvatori, T Lukasiewicz, T Xu, Z Bogacz, R Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title | Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title_full | Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title_fullStr | Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title_full_unstemmed | Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title_short | Inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| title_sort | inferring neural activity before plasticity as a foundation for learning beyond backpropagation |
| work_keys_str_mv | AT songy inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation AT millidgeb inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation AT salvatorit inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation AT lukasiewiczt inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation AT xuz inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation AT bogaczr inferringneuralactivitybeforeplasticityasafoundationforlearningbeyondbackpropagation |