CRCLLGApr 19, 2024

The Instruction Hierarchy: Training LLMs to Prioritize Privileged Instructions

arXiv:2404.13208v1359 citationsh-index: 18
Originality Incremental advance
AI Analysis

This addresses security vulnerabilities in LLMs for developers and users, though it is incremental as it builds on existing methods for instruction following.

The paper tackles the problem of LLMs being vulnerable to prompt injections and jailbreaks by proposing an instruction hierarchy to prioritize system prompts over untrusted user inputs, resulting in drastically increased robustness with minimal degradation in standard capabilities.

Today's LLMs are susceptible to prompt injections, jailbreaks, and other attacks that allow adversaries to overwrite a model's original instructions with their own malicious prompts. In this work, we argue that one of the primary vulnerabilities underlying these attacks is that LLMs often consider system prompts (e.g., text from an application developer) to be the same priority as text from untrusted users and third parties. To address this, we propose an instruction hierarchy that explicitly defines how models should behave when instructions of different priorities conflict. We then propose a data generation method to demonstrate this hierarchical instruction following behavior, which teaches LLMs to selectively ignore lower-privileged instructions. We apply this method to GPT-3.5, showing that it drastically increases robustness -- even for attack types not seen during training -- while imposing minimal degradations on standard capabilities.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes