Tamper-Resistant Safeguards for Open-Weight LLMs

TL;DR

This paper introduces TAR, a novel method to embed tamper-resistant safeguards into open-weight LLMs, making it difficult for adversaries to remove safety measures even after extensive fine-tuning, thereby enhancing security.

Contribution

We propose TAR, a new approach that significantly improves the tamper-resistance of safeguards in open-weight LLMs, addressing vulnerabilities of existing methods.

Findings

TAR greatly enhances tamper-resistance of safeguards.

TAR preserves the benign capabilities of LLMs.

Extensive evaluations show TAR's robustness against fine-tuning attacks.

Abstract

Rapid advances in the capabilities of large language models (LLMs) have raised widespread concerns regarding their potential for malicious use. Open-weight LLMs present unique challenges, as existing safeguards lack robustness to tampering attacks that modify model weights. For example, recent works have demonstrated that refusal and unlearning safeguards can be trivially removed with a few steps of fine-tuning. These vulnerabilities necessitate new approaches for enabling the safe release of open-weight LLMs. We develop a method, called TAR, for building tamper-resistant safeguards into open-weight LLMs such that adversaries cannot remove the safeguards even after hundreds of steps of fine-tuning. In extensive evaluations and red teaming analyses, we find that our method greatly improves tamper-resistance while preserving benign capabilities. Our results demonstrate that progress on tamper-resistance is possible, opening up a promising new avenue to improve the safety and security of open-weight LLMs.