Challenges can be seen as a type of game that motivates participants to solve serious tasks. As a result, competition organizers must develop effective game rules. However, these rules have multiple objectives beyond making the game enjoyable for participants. These objectives may include solving real-world problems, advancing scientific or technical areas, making scientific discoveries, and educating the public. In many ways, creating a challenge is similar to launching a product. It requires the same level of excitement and rigorous testing, and the goal is to attract ''customers'' in the form of participants. The process begins with a solid plan, such as a competition proposal that will eventually be submitted to an international conference and subjected to peer review. Although peer review does not guarantee quality, it does force organizers to consider the impact of their challenge, identify potential oversights, and generally improve its quality. This chapter provides guidelines for creating a strong plan for a challenge. The material draws on the preparation guidelines from organizations such as Kaggle 1 , ChaLearn 2 and Tailor 3 , as well as the NeurIPS proposal template, which some of the authors contributed to.

The generative AI revolution in recent years has been spurred by an expansion in compute power and data quantity, which together enable extensive pre-training of powerful text-to-image (T2I) models. With their greater capabilities to generate realistic and creative content, these T2I models like DALL-E, MidJourney, Imagen or Stable Diffusion are reaching ever wider audiences. Any unsafe behaviors inherited from pretraining on uncurated internet-scraped datasets thus have the potential to cause wide-reaching harm, for example, through generated images which are violent, sexually explicit, or contain biased and derogatory stereotypes. Despite this risk of harm, we lack systematic and structured evaluation datasets to scrutinize model behavior, especially adversarial attacks that bypass existing safety filters. A typical bottleneck in safety evaluation is achieving a wide coverage of different types of challenging examples in the evaluation set, i.e., identifying 'unknown unknowns' or long-tail problems. To address this need, we introduce the Adversarial Nibbler challenge. The goal of this challenge is to crowdsource a diverse set of failure modes and reward challenge participants for successfully finding safety vulnerabilities in current state-of-the-art T2I models. Ultimately, we aim to provide greater awareness of these issues and assist developers in improving the future safety and reliability of generative AI models. Adversarial Nibbler is a data-centric challenge, part of the DataPerf challenge suite, organized and supported by Kaggle and MLCommons.