tag:blogger.com,1999:blog-21605329.post4464785885564832892..comments2024-03-14T23:52:09.893-07:00Comments on The Neurocritic: The Big Ideas in Cognitive Neuroscience, ExplainedThe Neurocritichttp://www.blogger.com/profile/08010555869208208621noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-21605329.post-43205469830168930082017-05-18T16:55:04.690-07:002017-05-18T16:55:04.690-07:00Relished your reading list headers, especially &qu...Relished your reading list headers, especially "Everyone is wrong". The human condition. For a somewhat different take on wrongness, and on the relationship between language and vision, you might glance at this online blog/book:<br /><br />rewiring-neuroscience.com<br /><br />Regards, John HarrisJohn Harrishttp://rewiring-neuroscience.comnoreply@blogger.comtag:blogger.com,1999:blog-21605329.post-16717393011565893482017-04-20T02:44:09.007-07:002017-04-20T02:44:09.007-07:00Our claim that learning in cerebellar Purkinje cel...Our claim that learning in cerebellar Purkinje cells involves something quite different from LTD/LTP (Johansson et al., PNAS 2014) is hardly overturned by an an appeal to majority opinion or "consensus". The authors of the Poo et al paper (BMC Biology, 2016) did not comment our paper and had probably not read it when they wrote their review. In fact, they do not discuss the cerebellum at all. Our claim has also received further support in a more recent paper (Johansson et al., Cell Reports, 2015). It is also consistent with motor learning in mice which have had their LTD mechanism knocked out (Schonewille et al., Neuron 2011).Anonymoushttps://www.blogger.com/profile/05687171902068741864noreply@blogger.comtag:blogger.com,1999:blog-21605329.post-71821091467987946252017-04-19T09:58:53.589-07:002017-04-19T09:58:53.589-07:00One major problem with cost functions is the fact ...One major problem with cost functions is the fact that given a behavior or neural pattern, you can always cook up a cost function for which the observed data (or your preferred normative story) constitutes an optimal solution. So unless the cost function idea is further constrained by data, the idea of optimizing cost functions in neuroscience amounts to 'animals try to improve things'. Which no one would actually argue against.Yohanhttp://neurologism.comnoreply@blogger.com