Jailbreaking Large Language Models via Multi-Task Embedding-based Prompt
Open Access
- 12 April 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 139-143
Abstract
With the widespread application of large language models (LLMs), jailbreak attacks that bypass their security mechanisms have become an important threat. Existing methods suffer from limited effectiveness, as they use only one attack pattern at a time, along with excessive query costs. To address these limitations, this paper proposes the Multi-Task Embedding-based Attack (MTEA), a novel jailbreak approach that embeds malicious instructions within three different tasks (e.g., Code Understanding, Language Translation, and Pattern Adherence). By exploiting LLMs’ reduced performance when handling multiple tasks concurrently, MTEA disrupts their safety alignment mechanisms. Extensive experiments on six widely used LLMs (including GPT-4o and Gemini-2.5-pro) using the AdvBench benchmark show that MTEA achieves 100% attack success rate (ASR) and 100% following rate (FR) across all models. It outperforms state-of-the-art baselines by 48.0 ∼ 49.7% in ASR and 60.4 ∼ 60.7% in FR. Even against advanced defense mechanisms (Perplexity Filter and SmoothLLM), MTEA maintains 100% ASR, while reducing query costs by 90% compared to baselines. These results demonstrate the effectiveness and efficiency of MTEA.Keywords
This publication has 12 references indexed in Scilit:
- Hybrid Automated Program Repair by Combining Large Language Models and Program AnalysisACM Transactions on Software Engineering and Methodology, 2025
- TestLoter: A logic-driven framework for automated unit test generation and error repair using large language modelsJournal of Computer Languages, 2025
- Usage of Large Language Model for Code Generation Tasks: A ReviewSN Computer Science, 2025
- Evaluating the Generalizability of LLMs in Automated Program Repair2025
- Jailbreaking Black Box Large Language Models in Twenty Queries2025
- Exploring Limitations of LLM Capabilities with Multi-Problem Evaluation2025
- Using an LLM to Help With Code Understanding2024
- Play Guessing Game with LLM: Indirect Jailbreak Attack with Implicit Clues2024
- InferFix: End-to-End Program Repair with LLMs2023
- Not All Languages Are Created Equal in LLMs: Improving Multilingual Capability by Cross-Lingual-Thought Prompting2023