Versatile Weight Attack via Flipping Limited Bits

TL;DR

This paper introduces a novel, versatile bit-flip based weight attack on deep neural networks during deployment, formulated as an optimization problem and solved efficiently, demonstrating high attack effectiveness and stealthiness.

Contribution

It proposes a general formulation for weight attacks via limited bit flips, with specific cases for single sample and triggered sample attacks, solved through continuous optimization techniques.

Findings

Effective attack success demonstrated in experiments

Optimization-based approach outperforms heuristic methods

Applicable to various attack scenarios and purposes

Abstract

To explore the vulnerability of deep neural networks (DNNs), many attack paradigms have been well studied, such as the poisoning-based backdoor attack in the training stage and the adversarial attack in the inference stage. In this paper, we study a novel attack paradigm, which modifies model parameters in the deployment stage. Considering the effectiveness and stealthiness goals, we provide a general formulation to perform the bit-flip based weight attack, where the effectiveness term could be customized depending on the attacker's purpose. Furthermore, we present two cases of the general formulation with different malicious purposes, i.e., single sample attack (SSA) and triggered samples attack (TSA). To this end, we formulate this problem as a mixed integer programming (MIP) to jointly determine the state of the binary bits (0 or 1) in the memory and learn the sample modification. Utilizing the latest technique in integer programming, we equivalently reformulate this MIP problem as a continuous optimization problem, which can be effectively and efficiently solved using the alternating direction method of multipliers (ADMM) method. Consequently, the flipped critical bits can be easily determined through optimization, rather than using a heuristic strategy. Extensive experiments demonstrate the superiority of SSA and TSA in attacking DNNs.