Attacking LLM Detectors with Homoglyph-Based Attacks

Join DevFest Scotland Online for an brilliant session on LLM detectors! 

As large language models (LLMs) become more and more skilled at writing human-like text, the ability to detect what they generate is critical. This session explores a novel attack vector, homoglyph-based attacks, that effectively bypasses state-of-the-art LLM detectors.

We'll begin by explaining the idea behind homoglyphs, characters that look similar but are encoded differently. You'll learn how these can be used to manipulate tokenization and evade detection systems. We'll cover the mechanisms of how homoglyphs alter text representation, discuss their impact on existing LLM detectors, and present a comprehensive evaluation of their effectiveness against various detection methods.

Join us for an engaging exploration of this emerging threat and to gain insight into how security researchers can stay ahead of evolving evasion techniques.

Agenda

---

Speaker

Aldan Creo - Accenture Labs (Technology Research Specialist)

Antiguo alumno de la ETSE (2019-2023), ahora trabajando en investigación en IA en Dublín :)

Host

Patty O’Callaghan - Charles River Laboratories (Tech Director | AI/ML GDE)

Patty is a Venezuelan living in Scotland. She has +20 years of experience in the tech industry and a background education in Computer Science and Artificial Intelligence. She is a Technical Director at Charles River Laboratories.
In her free time, Patty is a dedicated mentor and educator, leading a Code Club for 10-15 year-olds at her local school. Additionally, she's an organiser at Goo…

Complete your event RSVP here: https://gdg.community.dev/events/details/google-gdg-glasgow-presents-attacking-llm-detectors-with-homoglyph-based-attacks/.