Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Fang, Zhicheng, Zheng, Jingjie, Fu, Chenxu, Xu, Wei
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908867900735488
author Fang, Zhicheng
Zheng, Jingjie
Fu, Chenxu
Xu, Wei
author_facet Fang, Zhicheng
Zheng, Jingjie
Fu, Chenxu
Xu, Wei
contents Jailbreak techniques for large language models (LLMs) evolve faster than benchmarks, making robustness estimates stale and difficult to compare across papers due to drift in datasets, harnesses, and judging protocols. We introduce JAILBREAK FOUNDRY (JBF), a system that addresses this gap via a multi-agent workflow to translate jailbreak papers into executable modules for immediate evaluation within a unified harness. JBF features three core components: (i) JBF-LIB for shared contracts and reusable utilities; (ii) JBF-FORGE for the multi-agent paper-to-module translation; and (iii) JBF-EVAL for standardizing evaluations. Across 30 reproduced attacks, JBF achieves high fidelity with a mean (reproduced-reported) attack success rate (ASR) deviation of +0.26 percentage points. By leveraging shared infrastructure, JBF reduces attack-specific implementation code by nearly half relative to original repositories and achieves an 82.5% mean reused-code ratio. This system enables a standardized AdvBench evaluation of all 30 attacks across 10 victim models using a consistent GPT-4o judge. By automating both attack integration and standardized evaluation, JBF offers a scalable solution for creating living benchmarks that keep pace with the rapidly shifting security landscape.
format Preprint
id arxiv_https___arxiv_org_abs_2602_24009
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking
Fang, Zhicheng
Zheng, Jingjie
Fu, Chenxu
Xu, Wei
Cryptography and Security
Artificial Intelligence
Computation and Language
Machine Learning
Jailbreak techniques for large language models (LLMs) evolve faster than benchmarks, making robustness estimates stale and difficult to compare across papers due to drift in datasets, harnesses, and judging protocols. We introduce JAILBREAK FOUNDRY (JBF), a system that addresses this gap via a multi-agent workflow to translate jailbreak papers into executable modules for immediate evaluation within a unified harness. JBF features three core components: (i) JBF-LIB for shared contracts and reusable utilities; (ii) JBF-FORGE for the multi-agent paper-to-module translation; and (iii) JBF-EVAL for standardizing evaluations. Across 30 reproduced attacks, JBF achieves high fidelity with a mean (reproduced-reported) attack success rate (ASR) deviation of +0.26 percentage points. By leveraging shared infrastructure, JBF reduces attack-specific implementation code by nearly half relative to original repositories and achieves an 82.5% mean reused-code ratio. This system enables a standardized AdvBench evaluation of all 30 attacks across 10 victim models using a consistent GPT-4o judge. By automating both attack integration and standardized evaluation, JBF offers a scalable solution for creating living benchmarks that keep pace with the rapidly shifting security landscape.
title Jailbreak Foundry: From Papers to Runnable Attacks for Reproducible Benchmarking
topic Cryptography and Security
Artificial Intelligence
Computation and Language
Machine Learning
url https://arxiv.org/abs/2602.24009