A Holistic Approach to Undesired Content Detection in the Real World

A Holistic Approach to Undesired Content Detection in the Real World | OpenAI

Skip to main content

[](https://openai.com/)

* Research

* Products

* Business

* Developers

* Company

* Foundation(opens in a new window)

Log inTry ChatGPT(opens in a new window)

Try ChatGPT(opens in a new window)Login

OpenAI

June 20, 2024

Publication

A holistic approach to undesired content detection in the real world

Read paper(opens in a new window)

We present a holistic approach to building a robust and useful natural language classification system for real-world content moderation. The success of such a system relies on a chain of carefully designed and executed steps, including the design of content taxonomies and labeling instructions, data quality control, an active learning pipeline to capture rare events, and a variety of methods to make the model robust and to avoid overfitting. Our moderation system is trained to detect a broad set of categories of undesired content, including sexual content, hateful content, violence, self-harm, and harassment. This approach generalizes to a wide range of different content taxonomies and can be used to create high-quality content classifiers that outperform off-the-shelf models.

OpenAI

June 20, 2024

Authors

Todor Markov, Chong Zhang, Sandhini Agarwal, Tyna Eloundou, Teddy Lee, Steven Adler, Angela Jiang, Lilian Weng

Read paper(opens in a new window)

Loading…

We present a holistic approach to building a robust and useful natural language classification system for real-world content moderation. The success of such a system relies on a chain of carefully designed and executed steps, including the design of content taxonomies and labeling instructions, data quality control, an active learning pipeline to capture rare events, and a variety of methods to make the model robust and to avoid overfitting. Our moderation system is trained to detect a broad set of categories of undesired content, including sexual content, hateful content, violence, self-harm, and harassment. This approach generalizes to a wide range of different content taxonomies and can be used to create high-quality content classifiers that outperform off-the-shelf models.

* Language

* Ethics & Safety

Authors

Todor Markov, Chong Zhang, Sandhini Agarwal, Tyna Eloundou, Teddy Lee, Steven Adler, Angela Jiang, Lilian Weng