{"id":"doi:10.1371/journal.pcbi.1010628","title":"Sleep prevents catastrophic forgetting in spiking neural networks by forming joint synaptic weight representations","date":"2020-01-01","section":"NeuroAI","quality":0.107723,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/","source_url":"https://doi.org/10.1371/journal.pcbi.1010628","text_available":true,"authors":["Ryan Golden","Jean Erik Delanois","Pavel Sanda","Maxim Bazhenov"],"outgoing":[{"source":"doi:10.1371/journal.pcbi.1010628","target":"arxiv:1708.02072","evidence":"title","id":"arxiv:1708.02072","title":"Measuring Catastrophic Forgetting in Neural Networks","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/358d4d451093ba5234c459f6323da4968d71d8be905b6612f5bae0162e9145f2/"},{"source":"doi:10.1371/journal.pcbi.1010628","target":"doi:10.1073/pnas.1611835114","evidence":"id","id":"doi:10.1073/pnas.1611835114","title":"Overcoming catastrophic forgetting in neural networks","url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/849082368a44d06b93ef3ba95bcfddc6672b7a5e8064ad2b2d19601639266e8b/"}],"incoming":[],"text_parts":[{"start":0,"end":14753,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html"},{"start":14753,"end":30634,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-2.html"},{"start":30634,"end":45876,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-3.html"},{"start":45876,"end":61189,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html"},{"start":61189,"end":65199,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html"}],"contents":[{"title":"Sleep prevents catastrophic forgetting in spiking neural networks by forming joint synaptic weight representations","offset":0,"level":1,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Abstract","offset":722,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Introduction","offset":1825,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Results","offset":5881,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Complementary complex foraging tasks can be robustly learned","offset":5893,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Catastrophic forgetting occurs following sequential but not interleaved training","offset":14117,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-1.html#text"},{"title":"Periods of sleep allow for sequential training without catastrophic forgetting","offset":19891,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-2.html#text"},{"title":"Receptive fields of decision-making neurons after sleep represent multiple tasks","offset":28091,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-2.html#text"},{"title":"Periods of sleep allow reintegration of new task without interference through renormalization of task-relevant synapses","offset":30509,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-2.html#text"},{"title":"Periods of sleep push the network towards the intersection of the solution manifolds representing Task 1 and Task 2 specific weight configurations","offset":35801,"level":3,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-3.html#text"},{"title":"Discussion","offset":41269,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-3.html#text"},{"title":"Methods","offset":49207,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Environment","offset":49219,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Network structure","offset":50463,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Policy","offset":52109,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Neuron models","offset":54409,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Synaptic plasticity","offset":56452,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Simulated Sleep","offset":59907,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"Support Vector Machine Training","offset":61151,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-4.html#text"},{"title":"2-D Synaptic Weight distributions (Figure 6)","offset":62114,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html#text"},{"title":"Distance from Solution Manifolds (Figure 7)","offset":63363,"level":4,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html#text"},{"title":"Author Contributions","offset":64654,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html#text"},{"title":"Conflicts of Interest","offset":65018,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html#text"},{"title":"Acknowledgements","offset":65051,"level":2,"url":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/part-5.html#text"}],"text_code_points":65199,"source_markdown":"https://vinnik-dmitry07.github.io/favorite-papers/agents/papers/7e34ad47e15cfd27cb9f93eef1399cfa51820e1382157b2c698c13548efc960a/source.md"}