Image credit:

The Music of Neuromatch

Image credit:

The Music of Neuromatch


Date
Event

The Music of Neuromatch

Neuromatch

Neuromatch Academy is a phenomenal organisation which has changed the way people think about teaching online that I’ve had the privilege to been a part of the last two summers. You can read about the inception and running of the Neuromatch Academy summer school in this very interesting and informative blog by blog by Gunnar Blohm. Or you can read about the dramatic last minute ban of Iranian students and TAs in 2020 due to some US sanctions and then even more dramatic laster minute lift of the ban. Or you can read the paper about how Neuromatch has championed global accessibility.

Neuromatch Academy 2021 was particularly significant for me as I created some of the course content and made the concept maps with the help of my 7 year old daughter. The process of writing material for Neuromatch is very edifying. I had the pleasure of discussing content materials with other content creators (Ella Batty, Arvind Kumar and Ulrik Beierholm) before and while writing the tutorials, as well as getting feedback from pre-pod teaching assistants before the tutorials were released. After all that there was a post-production team that made sure everything worked and edited and subtitled the videos. As someone who enjoys teaching, rarely have I had the time or resources to have a team to help develop the materials so this was a highlight of my teaching year and perhaps a high point of my 20+ years of teaching.

The Neuromatch materials are excellent and accessible to everyone across all time zones, so the next barrier to learning is ensuring that the students feel they can ask questions. A solution to this happened organically. Some of the volunteers were so invested in the materials they made playlists, tribute songs and comedy routines. None of this trivialized the content but were an Ode to the Joy of the materials. While these tributes might not directly impact the learning outcomes they created a more convivial learning atmosphere for the students, which can only be a good thing. Here, I will write about the link between music and Neuromatch and why I made playlists for the Neuromatch summer schools.

Music and Teaching

I began playing songs in lectures when due to COVID-19 we were forced to move to remote teaching. I felt that the start of every online lecture was very abrupt and to be honest I felt awkward as the only person with their camera on. Normally, with in person lectures I would begin by chatting about different things to help the students relax and encourage engagement. This was not possible in an online setting as for a multitude of different reasons most students could not turn on their cameras or microphones. I wanted to find a way to reduce both my own and their awkwardness and encourage engagement, so I decided to start every class with a song that had some link to the lecture content. The songs acted as a sound check for the students and as an ice-breaker and change of pace for everyone.

During the monotony and exhaustion of teaching online, picking songs that would have some relevance to the materials was both fun and challenging. The songs ranged from classical to pop to rock to jazz. For example, I picked the excellent “Zombie” by Fela Kuti to motivate a lecture on predator-prey models, and I was enjoying the song so much that I started dad chair-dancing along with my camera on, which I hope made everyone feel embarrassed but also feel a bit more at ease.

Other examples included picking “Sisyphus” by Andrew Bird to illustrate that most Numerical Methods are repeating the same thing again and again like rolling a rock up a hill. The most kudos I received for a song choice was when I played “The Boys In The Better Land” by Fontaines DC, which was not relevant to the lecture but the band had just been nominated for a Grammy. Overall, the student reaction was positive, maybe not about all the song choices, but they responded well to the rational, and it generated some chat and added variety to the day. Now to Neuromatch playlists.

The Neuromatch Playlists

As Neuromatch Academy in 2021 was massive with almost 5000 students from over 100 countries split between the computational neuroscience summer school and the deep learning summer school. Finding a link with the other members of neuromatch can seem daunting. The student engamement team did an excellent job organising events and even having a Neuromatch Academy's Got Talent which included musical, comedy, art and t-shirt design competitions. So to contribute in a small way I made two playlist one for each Neuromatch summer school with the lofty goals: Neuromatch Academy in 2021 had almost 5000 students from over 100 countries split between the Computational Neuroscience Summer School (June 30th-July 23rd, 2021) and the Deep Learning Summer School (August 2nd -August 23rd, 2021). So finding a link with the other members of Neuromatch and the materials could seem daunting. The student engagement team did an excellent job organising events such as a Neuromatch Academy’s got talent including a music, comedy, art and t-shirt design competitions. To contribute in a small way I made two playlists one for each Neuromatch Summer School with the lofty goals:
  • as a tribute the amazing course materials,
  • to add some levity to the learning experience,
  • to hopefully help some students make a connect to each other and the materials,
  • to encourage myself to look at all the materials,
  • lastly, as a thank you to all the people involved in organising Neuromatch Academy.

Computational Neuroscience Playlist

Computational Neuroscience Summer School started with a Python refresher which I paired it with “Holst: The Planets”. In particular the song “Jupyter: the Bringer of Joy”, which in my mind is the unofficial anthem of Computational Neuroscience Neuromatch as I believe Neuromatch is the Bringer of Joy to many students. To give a sample of the songs and the links, I picked “Don’t Give Up” Peter Gabriel & Kate Bush for the Dynamical Systems day to illustrate that the modelling of inhibition and excitation is like a beautiful duet between two equations. Also, this is the midpoint of the materials which can be a daunting moment for the students, so I figured the sentiment of the song seemed relevant:

🎶“Don’t give up now
We’re proud of who you are
Don’t give up
You know it’s never been easy.”🎶


Here is the complete tweet thread for choosing each song and my motivation for the Computational Neuroscience Summer School

and the corresponding Computational Neuroscience Summer School Spotify playlist:

Deep Learning Playlist

This is the first year of the Deep Learning summer school. So I figured the unofficial anthem would be the adapted song lyrics from "In Too Deep" by Sum 41 which might be a window into the minds of how some of the organisers and students: This is the first year of the Deep Learning Neuromatch Summer School. I started with “Rolling in the Deep” by Adele as the students were straight into the deep with the basics and Pytorch. On the Optimization day I paired “In too Deep” by Sum 41 which might have been a window into the minds of how some of the organisers and students were feeling:
🎶"'Cause I'm in too deep, and I'm trying to keep
Up above in my head, instead of going under"
Instead of going under I did neuromatch DL.🎶
Hopefully, this provided some reassurance to all students that there are moments when we all feel like we are in too deep when learning something new but have trust the process. Here is the full Deep Learning tweet thread with the details of the sometimes tenuous motivation of why I picked each song: and the corresponding Spotify Deep Learning Summer School playlist:

Musical Tribute

On the last day of the Deep Learning course, as part of the music competition, Megan Peters (Chairperson of Neuromatch) knocked the musical tributes to Neuromatch out of the park with her amazing song inspired by The Bee Gees “How Deep is Your Love?” and a small suggestion from Gunnar Blohm. Megan rewrote the song as “How Deep Can You Learn?”. Megan’s includes brilliant lyrics such as:
🎶“We’re living in a neural net,
optimizing….
When we all should
generalise!”🎶
It is a true musical and Deep Learning genius and I believe it should be the Neuromatch Deep Learning anthem.

Ode to Joy

For myself, making the playlists was enjoyable, I found some new music, and I read the course materials. I will be dipping back into them both again and again and again as they are excellent songs and materials. The songs and their links to the materials generated a bit of discussion on twitter and on the summer schools discord music channels, which was nice. Finally, I hope that when people listen to the playlists they know it as a grateful thank you from me to the students, the TAs, the volunteers and the board of Neuromatch.
Outreach
Avatar
John S Butler
Lecturer in Mathematics and Statistics

My research interests are the application of computational, statistical and numerical methods for basic and translational research in Neuroscience, Neurology and Optometry.