CBS Brain Science Seminar Series(CBSS)

Dr. Carsen Stringer

Date/Time

Thursday, June 5 2025, 9:00-10:30 (JST)

Abstract

Unsupervised pretraining in biological neural networks

Representation learning in neural networks may be implemented with supervised or unsupervised algorithms, distinguished by the availability of instruction. In sensory cortex, perceptual learning drives neural plasticity, but it is not known if this is due to supervised or unsupervised learning. Here we recorded populations of up to 90,000 neurons simultaneously from the primary visual cortex (V1) and higher visual areas (HVA), while mice learned multiple tasks as well as during unrewarded exposure to the same stimuli. Similar to previous studies, we found that neural changes in task mice were correlated with their behavioral learning. However, the neural changes were mostly replicated in mice with unrewarded exposure, suggesting that the changes were in fact due to unsupervised learning. The neural plasticity was highest in the medial HVAs and obeyed visual, rather than spatial, learning rules. Our neural results predict that unsupervised learning may accelerate subsequent task learning, a prediction which we validated with behavioral experiments. To investigate these unsupervised learning rules, we built simplified models of visual cortical responses. These models provided insight into how visual cortical circuits produce invariant representations, which may be used for visual learning and guiding behavior.

Registration

Registration Required.
CBS members should register at the CBSS.

  • Registration is valid for one year.
    You will receive an invitation mail from Zoom once you complete your registration.
    The zoom invitation mail will be sent 2-3 days before each seminar.
  • The registration deadline is NOON, three business days (JST) before each lecture date for first-time registrants.
    After the deadline, you will not be able to participate in the seminar.
  • Participation is limited only for researchers affiliated to research or academic institutions.
    Please provide an email address issued by your university/research institute.
    We cannot accept free email addresses such as gmail/hotmail/yahoo.

Be aware that your invitation e-mail for the CBSS may be incorrectly placed in your SPAM mailbox. If you cannot find the invitation mail, please check your SPAM mailbox, just in case. Thank you very much for your kind cooperation.

Register

For inquiries

cbss@ml.riken.jp