2026 AI Paper Reading Club #2

CG Marketing, 5/F, East Wing, 822 Lai Chi Kok Road, Cheung Sha Wan, Kowloon, 000000

GDG Hong Kong

AI Paper Reading Club is for ML engineers to explore AI research together. Join sessions to discuss "A Watermark for Large Language Models". Meet on October 6, 2026, from 7:30 to 8:45 pm at CG Marketing, Hong Kong. Perfect for those interested in AI and LLMs. Please read the papers in advance.

Oct 6, 11:30 AM – 1:00 PM (UTC)

27 RSVP'd

Key themes

About this event

**IMPORTANT: NEW ADDRESS FOR CG MARKETING: 5/F, East Wing, 822 Lai Chi Kok Road, Cheung Sha Wan (Lai Chi Kok MTR station D2 exit)

What is AI Paper Reading Club?

AI Paper Reading Club is a machine learning paper reading campaign with the sole purpose to facilitate the notion of “AI research” among ML engineers. As research in AI grows the importance by time, we wanted to encourage people to have a chance to read AI papers. With this campaign, AI communities will read, learn, and share the knowledge to train and nurture the future AI researchers.

Why Join Us?

Dive Deep into ML Research: Get firsthand exposure to groundbreaking papers and understand the mechanics behind the algorithms that are shaping the future.

Collaborative Learning: Share insights, ask questions, and engage in stimulating discussions with peers who share your passion for AI.

Bridge Theory and Practice: Learn how to translate complex research findings into practical applications and innovations.

How does it work?

These are book clubs in nature. The idea is going through latest AI papers sequentially together to have deeper understanding in AI research. What is additionally expected from the participants is to create resulting videos about those papers.

Upcoming Session (6/10/2026):

  • A Watermark for Large Language Models (Paper: https://arxiv.org/pdf/2301.10226)

    This talk explores A Watermark for Large Language Models (Kirchenbauer et al., 2023), a foundational framework for identifying machine-generated text without modifying model weights or accessing API parameters. By leveraging a pseudo-random hash to softly bias next-token logits toward a dynamically chosen "green list," this method embeds an imperceptible signal into generated text that can be detected from spans as short as 25 tokens with provable statistical confidence — establishing a fast, open-source approach to auditing and safety in generative AI.

Event Details

  • 📅 Date & Time: 6th October 2026 (Tuesday). 7:30 pm to 8:45 pm

  • 📍 Venue: CG Marketing, 5/F, East Wing, 822 Lai Chi Kok Road, Cheung Sha Wan, Kowloon, Hong Kong (Lai Chi Kok MTR station D2 exit) [NOTE: new address of the same venue sponsor]

  • 🗣️ Language: English / Cantonese

  • 🤝 Who should join: Anyone with knowledge in AI/ML, especially in LLMs

How to Prepare:

Please read / skim through the paper before the meeting, so that you can understand some of the basic concepts before our meet-up, as well as come prepared with questions or points for discussion.

When

When

Tuesday, October 6, 2026
11:30 AM – 1:00 PM (UTC)

Speaker

  • Ray Hsu

    The University of Hong Kong

    Lecturer

Facilitator

  • Kenneth Lui

    GDG Hong Kong Organizer

Partner

CG Marketing logo

CG Marketing

Organizers

  • Frankie Wu

    GDGHK Organizer

  • Kenneth Lui

    GDGHK Organizer

  • Lydia So

    GDGHK Helper

  • Keith Chong

    GDGHK Helper

  • Acesley Chan

    GDGHK Helper

  • Nicholas Benedict

    GDGHK Helper