Episode 14: Florian Fiebig on Hebbian Learning Networks

Matthew Taylor • Community Manager

In this episode, host Matt Taylor chats with Numenta Visiting Research Scientist Florian Fiebig. Florian is a recent graduate from the KTH Royal Institute of technology in Stockholm, Sweden with a PhD in computational neuroscience. His PhD thesis focuses on Hebbian memory networks and he regularly presents his work at Numenta research meetings. Florian’s thesis is titled, “Active Memory Processing on Multiple Timescales in Simulated Critical Networks with Hebbian Plasticity.”

Show Notes

  • 1:05    Intro to Florian
  • 2:41    Florian’s background and what led him to Numenta
  • 3:06    Continuous learning
  • 9:30     Does deep learning have anything similar to Hebbian learning?
  • 11:36   Different types of plasticity in Hebbian learning
  • 11:55   Long-term Potentiation (LTP)
  • 14:38   “So it turns out: Short-term potentiation is not always short-term potentiation”
  • 15:47   Two fast forms of plasticity: facilitation and augmentation
  • 17:57   Homeostatic mechanisms: the Bobcat example
  • 19:41   Let’s talk about working memory
  • 21:21   Associative nature of memory
  • 26:46   The brain as a massive filter
  • 28:16   Episodic memory vs. semantic memory
  • 30:38   Non-declarative memories
  • 32:47   How does the transfer process of initially acquired memory into something that is longer lasting work?
  • 35:05   The keys to remembering: repetition and relevance
  • 37:28   Attractors and dynamical systems
  • 44:15   The cortical attractor theory of neocortex or neocortical memory
  • 45:26   The binding problem
  • 55:00   Closing

Video

Download the full transcript of the podcast here.

 

Subscribe to Numenta On Intelligence: iTunes, Stitcher, Google Play, Spotify, RSS

Matthew Taylor • Community Manager