Tricking Retrievers with Influential Tokens: An Efficient Black-Box Corpus Poisoning Attack
This addresses a security problem for RAG systems by enabling more efficient adversarial attacks, though it is incremental as it builds on existing poisoning methods.
The paper tackles the vulnerability of retrieval-augmented generation (RAG) systems to corpus poisoning attacks by proposing DIGA, an efficient black-box method that achieves comparable or better attack success rates with significantly reduced time and memory usage.
Retrieval-augmented generation (RAG) systems enhance large language models by incorporating external knowledge, addressing issues like outdated internal knowledge and hallucination. However, their reliance on external knowledge bases makes them vulnerable to corpus poisoning attacks, where adversarial passages can be injected to manipulate retrieval results. Existing methods for crafting such passages, such as random token replacement or training inversion models, are often slow and computationally expensive, requiring either access to retriever's gradients or large computational resources. To address these limitations, we propose Dynamic Importance-Guided Genetic Algorithm (DIGA), an efficient black-box method that leverages two key properties of retrievers: insensitivity to token order and bias towards influential tokens. By focusing on these characteristics, DIGA dynamically adjusts its genetic operations to generate effective adversarial passages with significantly reduced time and memory usage. Our experimental evaluation shows that DIGA achieves superior efficiency and scalability compared to existing methods, while maintaining comparable or better attack success rates across multiple datasets.