Trojan Detection Through Pattern Recognition for Large Language Models

Abstract: Trojan backdoors can be injected into LLMs at various stages, including pretraining, fine-tuning, and in-context learning, posing a significant threat to the model's alignment. Due to the nature of causal language modeling, detecting these triggers is challenging given the vast search space. In this study, we propose a multistage framework for detecting Trojan triggers in LLMs consisting of token filtration, trigger identification, and trigger verification. We discuss existing trigger identification methods and propose two variants of a black-box trigger inversion method that rely on output logits, utilizing beam search and greedy decoding respectively. We show that the verification stage is critical in the process and propose semantic-preserving prompts and special perturbations to differentiate between actual Trojan triggers and other adversarial strings that display similar characteristics. The evaluation of our approach on the TrojAI and RLHF poisoned model datasets demonstrates promising results.