SPML: A DSL for Defending Language Models Against Prompt Attacks
This addresses the problem of securing application-specific chatbots against attacks for developers and organizations, though it is incremental as it builds on existing prompt defense research.
The paper tackles the vulnerability of LLM-based chatbots to malicious prompt attacks by introducing SPML, a domain-specific language that refines prompts and monitors inputs to prevent unethical applications and reduce costs, achieving superior performance over models like GPT-4 in experiments.
Large language models (LLMs) have profoundly transformed natural language applications, with a growing reliance on instruction-based definitions for designing chatbots. However, post-deployment the chatbot definitions are fixed and are vulnerable to attacks by malicious users, emphasizing the need to prevent unethical applications and financial losses. Existing studies explore user prompts' impact on LLM-based chatbots, yet practical methods to contain attacks on application-specific chatbots remain unexplored. This paper presents System Prompt Meta Language (SPML), a domain-specific language for refining prompts and monitoring the inputs to the LLM-based chatbots. SPML actively checks attack prompts, ensuring user inputs align with chatbot definitions to prevent malicious execution on the LLM backbone, optimizing costs. It also streamlines chatbot definition crafting with programming language capabilities, overcoming natural language design challenges. Additionally, we introduce a groundbreaking benchmark with 1.8k system prompts and 20k user inputs, offering the inaugural language and benchmark for chatbot definition evaluation. Experiments across datasets demonstrate SPML's proficiency in understanding attacker prompts, surpassing models like GPT-4, GPT-3.5, and LLAMA. Our data and codes are publicly available at: https://prompt-compiler.github.io/SPML/.