LLM Security and Safety: Insights from Homotopy-Inspired Prompt Obfuscation

TL;DR

This paper introduces a homotopy-inspired prompt obfuscation framework to analyze security vulnerabilities in Large Language Models, revealing weaknesses in safeguards and proposing a foundation for developing more robust safety measures.

Contribution

It presents a novel prompt obfuscation method inspired by homotopy theory to systematically study LLM security vulnerabilities and safety issues.

Findings

Identified weaknesses in current LLM safeguards.

Demonstrated influence of prompt engineering on model behavior.

Provided a framework for analyzing and mitigating vulnerabilities.

Abstract

In this study, we propose a homotopy-inspired prompt obfuscation framework to enhance understanding of security and safety vulnerabilities in Large Language Models (LLMs). By systematically applying carefully engineered prompts, we demonstrate how latent model behaviors can be influenced in unexpected ways. Our experiments encompassed 15,732 prompts, including 10,000 high-priority cases, across LLama, Deepseek, KIMI for code generation, and Claude to verify. The results reveal critical insights into current LLM safeguards, highlighting the need for more robust defense mechanisms, reliable detection strategies, and improved resilience. Importantly, this work provides a principled framework for analyzing and mitigating potential weaknesses, with the goal of advancing safe, responsible, and trustworthy AI technologies.