paperMay 2026Unreviewed

MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety

Jialin Song, Xiaodong Liu, Weiwei Yang, Wuyang Chen, Mingqian Feng, Xuekai Zhu, Jianfeng Gao

Abstract

We present MultiBreak, a scalable and diverse multi-turn jailbreak benchmark to evaluate large language model (LLM) safety. Multi-turn jailbreaks mimic natural conversational settings, making them easier to bypass safety-aligned LLM than single-turn jailbreaks. Existing multi-turn benchmarks are limited in size or rely heavily on templates, which restrict their diversity. To address this gap, we unify a wide range of harmful jailbreak intents, and introduce an active learning pipeline for expand

Framework mappings

OWASP Top 10 for LLM Applications

LLM01Prompt Injection

MITRE ATLAS

AML.T0054LLM Jailbreak

Suggested from the entry's categories.

Cite

@misc{song2026multibreak,
  title = {{MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety}},
  author = {Jialin Song and Xiaodong Liu and Weiwei Yang and Wuyang Chen and Mingqian Feng and Xuekai Zhu and Jianfeng Gao},
  year = {2026},
  month = may,
  eprint = {2605.01687},
  archivePrefix = {arXiv},
  url = {https://www.semanticscholar.org/paper/531a805a9f8b0646b82ef0181660b9d5967c6199}
}