Graph of Attacks with Pruning: Optimizing Stealthy Jailbreak Prompt Generation for Enhanced LLM Content Moderation

Conference proceeding

Open access

Daniel Schwartz, Dmitriy Bespalov, Zhe Wang, Ninad Kulkarni and Yanjun Qi

EMNLP 2025 - 2025 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Industry Track, pp 659-671

2025

Files and links (1)

url

https://doi.org/10.18653/v1/2025.emnlp-industry.46View

Published, Version of Record (VoR) Open

1 Record Views

Title: Graph of Attacks with Pruning: Optimizing Stealthy Jailbreak Prompt Generation for Enhanced LLM Content Moderation
Creators: Daniel Schwartz - Drexel University
Dmitriy Bespalov - Drexel University
Zhe Wang - Amazon Bedrock Science
Ninad Kulkarni - Amazon Bedrock Science
Yanjun Qi - University of Virginia
Publication Details: EMNLP 2025 - 2025 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Industry Track, pp 659-671
Number of pages: 13
Resource Type: Conference proceeding
Language: English
Academic Unit: Computer Science
Scopus ID: 2-s2.0-105039643703
Other Identifier: 991022194858204721