Immunization against harmful fine-tuning attacks

arXiv:2402.16382v1 Announce Type: new
Abstract: Approaches to aligning large language models (LLMs) with human values has focused on correcting misalignment that emerges from pretraining. However, this focus overlooks another source of misalignment: bad actors might purposely fine-tune LLMs to achieve harmful goals. In this paper, we present an emerging threat model that has arisen from alignment circumvention and fine-tuning attacks. However, lacking in previous works is a clear presentation of the conditions for effective defence. We propose a set of …

abstract actors arxiv attacks cs.cl fine-tuning focus human language language models large language large language models llms paper pretraining threat type values