2024 Global Challenge for Safe and Secure LLMs (Track 1) Prize Giving Ceremony

2024 Global Challenge for Safe and Secure LLMs (Track 1) Prize Giving Ceremony

16 October 2024 (Wednesday)
Time: 11.00am to 11.25am (GMT +8)

The 2024 Global Challenge for Safe and Secure LLMs, co-organized by AI Singapore and the CyberSG R&D Programme Office, focuses on addressing the vulnerabilities of Large Language Models (LLMs) to jailbreaking attacks. The Challenge takes place over two tracks from July 2024 to April 2025, with Track 1 having just concluded in September.

In this prize-giving ceremony, we celebrate the efforts of participants in Track 1, where teams designed prompts to expose LLM weaknesses, such as generating offensive or harmful content. Out of 111 teams, 32 surpassed the baseline, and the top 3 teams excelled in the final phase.

As we honour these achievements, the journey continues with Track 2, set for January 2025. This next phase will challenge participants to develop external defenses for LLMs that detect and block harmful outputs without modifying the model’s internal structure.