Supporting Human Raters with the Detection of Harmful Content using Large Language Models

June 18, 2024 Β· Declared Dead Β· πŸ› IEEE Symposium on Security and Privacy

πŸ‘» CAUSE OF DEATH: Ghosted
No code link whatsoever

"No code URL or promise found in abstract"

Evidence collected by the PWNC Scanner

Authors Kurt Thomas, Patrick Gage Kelley, David Tao, Sarah Meiklejohn, Owen Vallis, Shunwen Tan, Blaž Bratanič, Felipe Tiengo Ferreira, Vijay Kumar Eranti, Elie Bursztein arXiv ID 2406.12800 Category cs.CR: Cryptography & Security Citations 7 Venue IEEE Symposium on Security and Privacy Last Checked 3 months ago
Abstract
In this paper, we explore the feasibility of leveraging large language models (LLMs) to automate or otherwise assist human raters with identifying harmful content including hate speech, harassment, violent extremism, and election misinformation. Using a dataset of 50,000 comments, we demonstrate that LLMs can achieve 90% accuracy when compared to human verdicts. We explore how to best leverage these capabilities, proposing five design patterns that integrate LLMs with human rating, such as pre-filtering non-violative content, detecting potential errors in human rating, or surfacing critical context to support human rating. We outline how to support all of these design patterns using a single, optimized prompt. Beyond these synthetic experiments, we share how piloting our proposed techniques in a real-world review queue yielded a 41.5% improvement in optimizing available human rater capacity, and a 9--11% increase (absolute) in precision and recall for detecting violative content.
Community shame:
Not yet rated
Community Contributions

Found the code? Know the venue? Think something is wrong? Let us know!

πŸ“œ Similar Papers

In the same crypt β€” Cryptography & Security

Died the same way β€” πŸ‘» Ghosted