We’re a community dedicated to understanding the foundations of large language models through thoughtful, rigorous discussion. Each week, we gather in downtown Vancouver to engage with influential papers, exchange ideas, and learn together in a welcoming, intellectually curious environment.
Please read the paper beforehand (It's okay not to understand everything), meeting is primarily for discussion.
📖 This Week’s Paper
Efficient Memory Management for Large Language Model Serving with PagedAttention
arxiv.org/abs/2309.06180
🕒 Schedule
6:00 PM — Arrival, introductions, and announcements
6:30–7:30 PM — Paper presentation and discussion
7:30 PM onward — continued conversation
Past Readings: github.com/meghakalia/LLM_paper_reading_group
Discord : discord.gg/CSmN35Ebm
(By friends at a larger machine learning reading group)
Ever wondered why the web is the way it is — or who decides where it goes next? The W3C's Technical Architecture Group (TAG) is a group of 12 people tasked with stewardship of the web platform. They review proposals, advise spec authors, and write documents like the Design Principles that shape how new web technologies are built. The TAG is in Vancouver for their face-to-face meeting, and they're joining us for an open panel. Bring your questions about the platform, the opportunities it brings, the challenges it faces, and the features you'd like to see. They're keen to hear from developers and content authors — this is a conversation, not a lecture. Panelists Heather Flanagan — Spherical Cow Consulting Jeffrey Yasskin — Google Brian Kardell — Igalia Yves Lafon — W3C Christian Liebel — Thinktecture Schedule 5:30 — Doors open, food & mingling 6:00 — “What’s New in Web? 2026 Edition” (Christian Liebel) 6:30 — Panel (moderator: Marcos Cáceres, Apple) Free to attend. Everyone welcome — from folks who just started writing JavaScript to people who've been arguing about specs for a decade.
Every Thursday at DCTRL (328 West Hastings) For the latest information, see [Vancouver Crypto Events · Events Calendar](https://luma.com/yvr)