Cleaning missing values: The human-generated questions may be unanswerable. Thus, we remove the Q&A items that lack available answers. Additionally, documents lacking any valid Q&A pairs are also removed.

For more information, please refer to the Distribution and Maintenance subsections of the datasheet provided in J. The annotations are in the COCO format. We provide detailed explanations on how the dataset can be read in the code repository.

The dataset will be available for a minimum of five years, with no plans for removal. We will ensure ongoing maintenance to verify and maintain data accessibility. For what purpose was the dataset created? Was there a specific task in mind? Who created the dataset (e.g., which team, research group) and on behalf of which Who funded the creation of the dataset?

Warning: This paper contains data and model outputs which are offensive in nature. Institutional Review Board (IRB) and obtained an exempt decision. Additionally, potential bias may arise due to the high cultural specificity of human reviewers. For instance, "explicit sexual content" is defined as "including Each video was evaluated by at least three volunteers. Following the initial assessment, we conduct a secondary cross-validation.

The recent development of Sora leads to a new era in text-to-video (T2V) generation. Along with this comes the rising concern about its safety risks.

However, a pivotal aspect of alignment for honesty involves discerning an LLM's

While Large Language Models (LLMs) display versatile functionality, they continue to generate harmful, biased, and toxic content, as demonstrated by the prevalence of human-designed jailbreaks .