AI Heap
Published on

Alignment-Enhanced Decoding:Defending via Token-Level Adaptive Refining of Probability Distributions

arXiv:2408.07663 - [arXiv,PDF]
Authors
  • Name
    Quan Liu
  • Name
    Zhenhong Zhou
  • Name
    Longzhu He
  • Name
    Yi Liu
  • Name
    Wei Zhang
  • Name
    Sen Su
  • Affiliation
    Beijing University of Posts and Telecommunications
Large language models are susceptible to jailbreak attacks, which can result in the generation of harmful content. While prior defenses mitigate these risks by perturbing or inspecting inputs, they ignore competing objectives, the underlying cause of alignment failures. In this paper, we propose Alignment-Enhanced Decoding (AED), a novel defense that employs adaptive decoding to address the root causes of jailbreak issues. We first define the Competitive Index to quantify alignment failures and utilize feedback from self-evaluation to compute post-alignment logits. Then, AED adaptively combines AED and post-alignment logits with the original logits to obtain harmless and helpful distributions. Consequently, our method enhances safety alignment while maintaining helpfulness. We conduct experiments across five models and four common jailbreaks, with the results validating the effectiveness of our approach. Code is available at https://github.com/GIGABaozi/AED.git.