Workshop infra

.github/workflows/main.yml

@@ -2,7 +2,7 @@ name: Deploy Hugo site to Pages
 
 on:
   push:
-    branches: ["main", "cache-images"]
+    branches: ["main", "student-talk"]
 
 permissions:
   contents: read
@@ -63,7 +63,7 @@ jobs:
 
       - name: Generate OG Images
         run: npm run og-images
-        
+
       - name: Fetch GitHub Stars
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.gitignore

@@ -26,3 +26,4 @@ yarn.lock
 content/**/*-og-*.jpg
 static/images/og-image-*.jpg
 tmp/og-cache-manifest.json
+/data/github_stars.json

archetypes/hacking-hours.md

@@ -13,9 +13,8 @@ description: "Join us for a hands-on hacking hours on [Topic] with [Speaker Name
 upcoming: true # Set to 'true' for future events. Change to 'false' after the event.
 video: ""      # After the event, add the YouTube video ID (e.g., "dQw4w9WgXcQ").
 image: "hacking-hours-banner.png" # Main banner image for social sharing (1200x630px). Place in this folder.
-speaker_photo: "speaker-photo.jpg" # Speaker's photo. Place in this folder.
 type: "hacking-hours" # IMPORTANT: Do not change this line.
-speaker_bio: "A brief biography of the speaker. What makes them the right person to talk about this topic?"
+show_author_bios: true # Set to true to display author bios from contributor profiles.
 # Optional: Add the file name (slug) of any software pages this event is related to.
 # e.g., ["spyx", "snntorch"]
 software_tags: []

archetypes/student-talks.md

@@ -10,9 +10,8 @@ description: "Join us for an insightful student talk on [Topic] by [Speaker Name
 upcoming: true # Set to 'true' for future events. Change to 'false' after the event.
 video: ""      # After the event, add the YouTube video ID (e.g., "dQw4w9WgXcQ").
 image: "student-talk-banner.png" # Main banner image (1200x630px ideal for sharing). Place in this student talk's folder.
-speaker_photo: "speaker-photo.jpg" # Speaker's photo. Place in this student talk's folder.
 type: "student-talks" # IMPORTANT: Do not change this line for student talk events.
-speaker_bio: "A brief biography of the speaker. Highlight their expertise relevant to the student talk topic."
+show_author_bios: true # Set to true to display author bios from contributor profiles.
 # Optional: Add the file name (slug) of any software pages this event is related to.
 # e.g., ["spyx", "snntorch"]
 software_tags: []

archetypes/workshops.md

@@ -10,9 +10,8 @@ description: "Join us for an insightful workshop on [Topic] by [Speaker Name]. D
 upcoming: true # Set to 'true' for future events. Change to 'false' after the event.
 video: ""      # After the event, add the YouTube video ID (e.g., "dQw4w9WgXcQ").
 image: "workshop-banner.png" # Main banner image (1200x630px ideal for sharing). Place in this workshop's folder.
-speaker_photo: "speaker-photo.jpg" # Speaker's photo. Place in this workshop's folder.
 type: "workshops" # IMPORTANT: Do not change this line for workshop events.
-speaker_bio: "A brief biography of the speaker. Highlight their expertise relevant to the workshop topic."
+show_author_bios: true # Set to true to display author bios from contributor profiles.
 # Optional: Add the file name (slug) of any software pages this event is related to.
 # e.g., ["spyx", "snntorch"]
 software_tags: []

content/contributors/alexandre-marcireau/index.md

@@ -2,10 +2,17 @@
 title: "Alexandre Marcireau"
 description: "Software developer in neuromorphic computing, recognized for his contributions to the Faery event processing library for event-based camera data."
 social:
+  - icon: "fa fa-house"
+    link: "https://www.westernsydney.edu.au/marcs/about/our_people/researchers/dr_alexandre_marcireau"
+    title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/alexandre-marcireau-6a634b6b/"
+    title: "linkedin"
   - icon: "fa-brands fa-github"
     link: "https://github.com/aMarcireau"
     title: "github"
 draft: false
 image: alexandre-marciraeu.png
 ---
 Alexandre Marcireau is a software developer and contributor in the neuromorphic computing space, known for his work on the Faery event processing library.
+https://www.westernsydney.edu.au/marcs/about/our_people/researchers/dr_alexandre_marcireau

content/contributors/christian-pehle/index.md

@@ -0,0 +1,17 @@
+---
+title: "Christian Pehle"
+description: "Researcher at Cold Spring Harbor Laboratory. Maintainer of Norse and jaxsnn."
+image: "christian-pehle.png"
+social:
+  - icon: "fa fa-house"
+    link: "https://cpehle.github.io/"
+    title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/christian-pehle-37820a209"
+    title: "linkedin"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/cpehle"
+    title: "github"
+draft: false
+---
+Christian Pehle is a researcher at Cold Spring Harbor Laboratory. He is a maintainer of the Norse and jaxsnn software frameworks.

content/contributors/dylan-muir/index.md

@@ -2,7 +2,14 @@
 title: "Dylan Muir"
 description: "VP Global Research Operations at SynSense, specialist in neural computation architectures. Directs research vision and neural architecture development."
 image: "dylan-muir.png"
-social: []
+social:
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/dylanmuir/"
+    title: "linkedin"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/DylanMuir"
+    title: "github"
 draft: false
 ---
 Dylan Muir is the Vice President for Global Research Operations; Director for Algorithms and Applications; and Director for Global Business Development at SynSense. Dr. Muir is a specialist in architectures for neural computation. He has published extensively in computational and experimental neuroscience. At SynSense he is responsible for the company research vision, and directing development of neural architectures for signal processing. Dr. Muir holds a Doctor of Science (PhD) from ETH Zurich, and undergraduate degrees (Masters) in Electronic Engineering and in Computer Science from QUT, Australia.
+

content/contributors/gregor-lenz/index.md

@@ -6,6 +6,9 @@ social:
   - icon: "fa fa-house"
     link: "https://lenzgregor.com/"
     title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/gregorlenz/"
+    title: "linkedin"
   - icon: "fa-brands fa-github"
     link: "https://github.com/biphasic"
     title: "github"

content/contributors/hananel-hazan/index.md

@@ -0,0 +1,18 @@
+---
+title: "Hananel Hazan"
+description: "Research scientist at the Allen Discovery Center at Tufts University and a leading developer of the BindsNET framework."
+image: "hananel-hazan.jpg"
+social:
+  - icon: "fa fa-house"
+    link: "https://hananel.hazan.org.il/"
+    title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/hananel-hazan/"
+    title: "linkedin"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/hananel-hazan"
+    title: "github"
+draft: false
+---
+Hananel Hazan is a research scientist at the Allen Discovery Center at Tufts University and a leading developer of the BindsNET framework.
+

content/contributors/james-knight/index.md

@@ -0,0 +1,18 @@
+---
+title: "James Knight"
+description: "Senior Research Software Engineer at the University of Sussex, leading developer of the GeNN neural simulation environment."
+image: "james-knight.jpeg"
+social:
+  - icon: "fa fa-house"
+    link: "https://profiles.sussex.ac.uk/p415734-james-knight"
+    title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/james-knight-61833211b/"
+    title: "linkedin"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/neworderofjamie"
+    title: "github"
+draft: false
+---
+
+James Knight is a Senior Research Software Engineer at the University of Sussex and a leading developer of the GeNN neural simulation environment.

content/contributors/jens-e-pedersen/index.md

@@ -6,6 +6,9 @@ social:
   - icon: "fa fa-house"
     link: "https://jepedersen.dk/"
     title: "website"
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/jens-egholm-pedersen-69543117/"
+    title: "linkedin"
   - icon: "fa-brands fa-github"
     link: "https://github.com/jegp"
     title: "github"

content/contributors/marcel-stimberg/index.md

@@ -0,0 +1,14 @@
+---
+title: "Marcel Stimberg"
+description: "Research engineer at Sorbonne Université. Leading developer of the Brian neural simulator."
+image: "marcel-stimberg.webp"
+social:
+  - icon: "fa fa-house"
+    link: "https://marcel.stimberg.info/"
+    title: "website"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/mstimberg"
+    title: "github"
+draft: false
+---
+Marcel Stimberg is a research engineer at Sorbonne Université and a leading developer of the Brian neural simulator.

content/contributors/matei-stan/index.md

@@ -0,0 +1,16 @@
+---
+title: "Matei Stan"
+description: "Matei Stan is a PhD student at the University of Manchester, focusing on State Space Models (SSMs) and their application in Spiking Neural Networks for long-range sequential tasks."
+image: "matei-stan.jpg"
+social:
+  - icon: "fa-brands fa-linkedin"
+    link: "https://www.linkedin.com/in/matei-stan"
+    title: "linkedin"
+draft: false
+---
+
+Matei Stan is a third-year PhD student in the Department of Computer Science at the University of
+Manchester, UK. He is supervised by Dr Oliver Rhodes in the Advanced Processor Technologies
+(APT) group. In his PhD work, Matei has primarily focused on the applications of deep State Space
+Models (SSMs), such as S4, in neuromorphic computing, and their potential in scaling
+energy-efficient algorithms for long-range sequential tasks.

content/contributors/terry-stewart/index.md

@@ -0,0 +1,15 @@
+---
+title: "Terry Stewart"
+description: "Research scientist focused on learning, memory, and biologically plausible SNNs. Lead developer of the Nengo neural simulator."
+image: "terry-stewart.png"
+social:
+  - icon: "fa fa-house"
+    link: "https://compneuro.uwaterloo.ca/people/terrence-c-stewart.html"
+    title: "website"
+  - icon: "fa-brands fa-github"
+    link: "https://github.com/tcstewar"
+    title: "github"
+draft: false
+---
+Terry Stewart is a research scientist focused on learning, memory, and biologically plausible SNNs. He is the lead developer of the Nengo neural simulator.
+

content/neuromorphic-computing/student-talks/learning-long-sequences-in-snns/index.md

@@ -0,0 +1,31 @@
+---
+title: "Student Talk: Learning Long Sequences in Spiking Neural Networks with Matei Stan"
+author:
+  - "Matei Stan"
+date: 2025-07-27
+start_time: "08:30"
+end_time: "09:45"
+time_zone: "EST"
+description: "Explore how State Space Models (SSMs) combined with Spiking Neural Networks (SNNs) can outperform Transformers on long-sequence tasks, and learn about a novel feature mixing layer that challenges assumptions about binary activations."
+upcoming: true
+upcoming_url: "https://teams.microsoft.com/l/meetup-join/19%3Ameeting_OTBkNTY5MjgtMjE3Ni00OTFmLWEwNzktN2QwZTU1NWIxNDc2%40thread.v2/0?context=%7B%22Tid%22%3A%22c152cb07-614e-4abb-818a-f035cfa91a77%22%2C%22Oid%22%3A%223f444780-d657-4917-993e-0f42adeff90e%22%7D"
+video: ""
+image: "banner.png"
+speaker_photo: "matei-stan.jpg"
+type: "student-talks"
+speaker_bio: "Matei Stan is a third-year PhD student in the Department of Computer Science at the University of Manchester, UK. He is supervised by Dr Oliver Rhodes in the Advanced Processor Technologies (APT) group. In his PhD work, Matei has primarily focused on the applications of deep State Space Models (SSMs), such as S4, in neuromorphic computing, and their potential in scaling energy-efficient algorithms for long-range sequential tasks."
+---
+
+Matei’s published work, “Learning Long Sequences in Spiking Neural Networks” [1], systematically investigates, for the first time, the intersection of the State‑of‑The‑Art State Space Models (SSMs) with Spiking Neural Networks (SNNs) for long‑range sequence modelling. Results suggest that SSM‑based SNNs can outperform the Transformer on all tasks of a well‑established long‑range sequence modelling benchmark - the “Long-Range Arena” [2]. It is also shown that the SSM‑based SNNs can outperform current State‑of‑The‑Art SNNs with fewer parameters on sequential image classification. Finally, a novel feature mixing layer is introduced, improving SNN accuracy while challenging assumptions about the role of binary activations in SNNs. This work paves the way for deploying powerful SSM-based architectures, such as Large Language Models, on neuromorphic hardware for energy-efficient long-range sequence modelling.
+
+This talk will highlight, at a high level, the similarities in computational primitives between SSMs and the existing neuromorphic standards such as Leaky Integrate-and-Fire (LIF) neurons. It will also focus on the specific drawbacks brought about by the introduction of binary activations in SSMs, as well as the extent to which these can be mitigated by the development of more accurate surrogate gradient methods that account for non-differentiability. Finally, arguments will be presented in favour of separating biological plausibility from energy efficiency in attempting to create scalable neuromorphic solutions.
+
+**Contents of the talk:**
+- Introduction to long-range sequence modelling
+- Link between SSMs and SNNs
+- Binary S4D and the Gated Spiking Unit
+- Q&A
+
+**References**
+[1]: Stan, M.I. and Rhodes, O., 2024. Learning long sequences in spiking neural networks. Scientific Reports, 14(1), p.21957.
+[2]: Tay, Y., Dehghani, M., Abnar, S., Shen, Y., Bahri, D., Pham, P., Rao, J., Yang, L., Ruder, S. and Metzler, D., 2020. Long range arena: A benchmark for efficient transformers. arXiv preprint arXiv:2011.04006.